CN117953016B - Flood discharge building exit area slope dangerous rock monitoring method and system - Google Patents
Flood discharge building exit area slope dangerous rock monitoring method and system Download PDFInfo
- Publication number
- CN117953016B CN117953016B CN202410356023.XA CN202410356023A CN117953016B CN 117953016 B CN117953016 B CN 117953016B CN 202410356023 A CN202410356023 A CN 202410356023A CN 117953016 B CN117953016 B CN 117953016B
- Authority
- CN
- China
- Prior art keywords
- dangerous rock
- video image
- target
- tracking
- side slope
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000011435 rock Substances 0.000 title claims abstract description 214
- 238000012544 monitoring process Methods 0.000 title claims abstract description 125
- 238000000034 method Methods 0.000 title claims abstract description 108
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 118
- 238000001514 detection method Methods 0.000 claims abstract description 63
- 230000008569 process Effects 0.000 claims description 32
- 230000002159 abnormal effect Effects 0.000 claims description 30
- 238000012545 processing Methods 0.000 claims description 29
- 230000004927 fusion Effects 0.000 claims description 25
- 230000033001 locomotion Effects 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 12
- 230000006872 improvement Effects 0.000 claims description 12
- 230000009191 jumping Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 10
- 230000003044 adaptive effect Effects 0.000 claims description 9
- 238000011160 research Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 8
- 238000010191 image analysis Methods 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 claims description 7
- 238000003708 edge detection Methods 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000010365 information processing Effects 0.000 abstract description 2
- 230000008859 change Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 206010063385 Intellectualisation Diseases 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000000917 particle-image velocimetry Methods 0.000 description 2
- 230000006798 recombination Effects 0.000 description 2
- 238000005215 recombination Methods 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Image Analysis (AREA)
- Alarm Systems (AREA)
Abstract
The invention belongs to the technical field of hydraulic engineering information processing, and discloses a method and a system for monitoring slope dangerous rock in an outlet area of a flood discharge building, wherein the method comprises the steps of distributing video monitoring instruments in the interior and the exterior of the slope dangerous rock, collecting monitoring data of the slope dangerous rock in real time, and generating a monitoring image of the slope dangerous rock; acquiring a real-time monitoring image, extracting a moving target area from the dangerous rock on the side slope based on a moving target detection algorithm, and completing the real-time monitoring of the dangerous rock on the side slope; analyzing each frame of real-time monitoring image of the side slope dangerous rock by taking the real-time monitoring image time factor of the side slope dangerous rock as input, and carrying out real-time and accurate tracking and positioning on the side slope dangerous rock based on a moving target tracking algorithm of an improved continuous self-adaptive mean tracking Camshift algorithm; analyzing and pre-judging the stability of the side slope dangerous rock through a prediction model, determining whether the current state of the side slope dangerous rock is normal, and predicting and pre-warning. The invention can early warn potential side slope disaster risks in time.
Description
Technical Field
The invention relates to the technical field of hydraulic engineering information processing, in particular to a method and a system for monitoring slope dangerous rock in an outlet area of a flood discharge building.
Background
In the areas of China, the mountain areas are larger and account for about two thirds of the areas of China, and geological disasters frequently occur due to the fact that the topography and geological conditions of the mountain areas are complicated, wherein dangerous rock collapse, landslide, debris flow and the like are main geological disaster types. In recent years, with the massive construction of hydraulic engineering, a large number of artificial slopes are generated, so that accidents caused by dangerous rock collapse of the slopes are gradually increased, huge losses are brought to people in the country, and the stable development of society is restricted. After secondary excavation of the rock slope, dangerous rock collapse of the slope can be caused artificially, so that huge potential safety hazards are brought to constructors and related facility equipment, a large amount of rocks form accumulation after the dangerous rock collapse of the slope, traffic is blocked in a form of a damming body, and engineering is affected; in order to reduce life and property loss caused by dangerous rock collapse of a side slope, geological disaster management work needs to be continuously enhanced, and meanwhile, high-risk areas where disasters occur are monitored through a series of scientific and technical means, so that people are helped to avoid risks, and loss caused by dangerous rock collapse of the side slope is reduced to the minimum. Therefore, the method has great practical significance in the aspect of protecting the life and property safety of people and the construction of national infrastructure when being used for monitoring the side slope dangerous rock area in real time, and has great economic value. However, the existing hydraulic engineering slope engineering monitoring system is difficult to realize real-time monitoring and prejudging, so that the hydraulic engineering slope engineering has great potential safety hazards, and the safety of constructors and facility equipment is not facilitated.
First, application number: CN202110455107.5 discloses a high-precision non-contact slope dangerous rock monitoring and early warning method, which specifically comprises the following steps: laying a monitoring instrument; calibrating a camera; starting a vibration measuring module to acquire the speed time course of the slope dangerous rock surface; starting a large-scale particle image velocimetry module to acquire a slope dangerous rock image; the calculation module calculates the vibration dominant frequency and the displacement change rate of the side slope dangerous rock according to the speed time course and the side slope dangerous rock image; monitoring vibration dominant frequency and displacement change rate of the side slope dangerous rock in real time, sending out early warning when the vibration dominant frequency and displacement change rate reach a preset threshold value, and adjusting shooting frequency of the large-scale particle image velocimetry module according to early warning information; although the remote, non-contact and real-time monitoring and early warning of the side slope dangerous rock are realized, the method is superior to the traditional side slope dangerous rock monitoring method, and has good application prospect; however, detection of moving slope images cannot be realized in a complex scene, so that the recognition accuracy of slope dangerous rock is low.
Second prior art, application number: CN202211371060.5 discloses a slope dangerous rock identifying and monitoring method based on unmanned aerial vehicle inspection, comprising the following steps: step 1: acquiring slope image information by adopting unmanned aerial vehicle inspection to obtain a first point cloud model, and performing vegetation removal pretreatment on the first point cloud model; step 2: extracting a main joint normal vector direction of the side slope based on the first point cloud model; step 3: determining the dangerous rock position of the side slope based on the direction of the normal vector of the main joint of the side slope by using a bare projection analysis method and a normal vector matching method; step 4: and (3) repeating the steps 1 to 3 for the same slope according to a preset time interval to obtain a second point cloud model, and carrying out point cloud model registration and relative distance calculation on the second point cloud model and the first point cloud model to obtain a slope change result. Although the method can solve the problem of insufficient contactless rapid recognition and research of the dangerous rock of the side slope and realize the technical effect of subsequent monitoring of the change of the side slope; however, unmanned aerial vehicles are high in cost, limited in cruising time and range and poor in real-time performance.
Third, application number: CN202210033838.5 discloses a slope dangerous rock collapse early warning method and device based on the characteristics of inclination and strong vibration, comprising: acquiring geometric information of a rock mass main control fracture, and acquiring rock mass main control fracture expansion characteristics and rock mass block structure characteristics based on the geometric information; based on the main control crack expansion characteristic and the blocking structural characteristic, arranging a micro-electromechanical system MEMS inclination vibration sensor; based on the MEMS inclination angle vibration sensor, the trend of the accumulated inclination angle variation, the range of the inclination angle variation speed and the state of the strong vibration acceleration signal are obtained; and judging the early warning level of each rock mass in the rock mass block structure based on the trend of the accumulated inclination angle variation, the range of the inclination angle variation speed and the state of the strong vibration acceleration signal. Although the method has the advantages of simple operation, low cost, low power consumption, reliable early warning judgment result and the like, the method can be widely applied to rock mass instability early warning of engineering slopes such as mountain town slopes, highways, railways and the like; but the function is single, only the early warning function is realized, and the real-time tracking and positioning of the target image of the side slope dangerous rock are lacked, so that the early warning result has errors.
The first, second and third existing technologies have the problems of single target monitoring function of slope dangerous rock and poor real-time tracking and positioning of target images, and larger early warning result errors; therefore, the invention provides the method and the system for monitoring the slope dangerous rock in the exit area of the flood discharge building, the detection algorithm can detect the moving target in a complex scene, has stronger adaptability to the target speed, and meets the requirement of the actual environment of the slope dangerous rock on the detection algorithm; the moving target tracking algorithm based on the improved Camshift is adopted, the integrity and the stability of the tracking algorithm are improved, and real-time and accurate tracking and positioning of the side slope dangerous rock are realized.
Disclosure of Invention
The invention mainly aims to provide a method and a system for monitoring slope dangerous rock in an exit area of a flood discharge building, which are used for solving the problems of relatively single target monitoring function of the slope dangerous rock, insufficient intelligent degree, relatively poor real-time tracking and positioning of target images and relatively large early warning result error in the prior art.
In order to achieve the above purpose, the present invention provides the following technical solutions:
A monitoring method for slope dangerous rock in the exit area of a flood discharge building specifically comprises the following steps:
distributing video monitoring instruments in the inner part and the outer part of the side slope dangerous rock, collecting monitoring data of the side slope dangerous rock in real time, and generating a monitoring image of the side slope dangerous rock;
Acquiring a real-time monitoring image, extracting a moving target area from the dangerous rock of the side slope based on a moving target detection algorithm fused by self-adaptive mixed color space information, and completing the real-time monitoring of the dangerous rock of the side slope;
Analyzing each frame of real-time monitoring image of the side slope dangerous rock by taking the real-time monitoring image time factor of the side slope dangerous rock as input, and carrying out real-time and accurate tracking and positioning on the side slope dangerous rock based on a moving target tracking algorithm of an improved continuous self-adaptive mean tracking Camshift algorithm;
Analyzing and pre-judging the stability of the side slope dangerous rock through a prediction model, determining whether the current state of the side slope dangerous rock is normal, and predicting and pre-warning.
As a further improvement of the present invention, collecting monitoring data of a slope dangerous rock in real time and generating a monitoring image of the slope dangerous rock includes:
A hardware platform of the side slope dangerous rock monitoring system is built through the mutual matching of the video image acquisition unit, the video image processing unit and the video image display unit;
a real-time video image sequence of the side slope dangerous rock area is acquired in real time through a video image acquisition unit; the video image processing unit analyzes and judges the video image through a moving target detection algorithm and a moving target tracking algorithm, when the abnormal state of the slope dangerous rock occurs, the video image processing unit timely acquires the image information of the slope dangerous rock movement, sends out an alarm and stores the processed field picture so as to be convenient for the call of related departments;
The video image display unit visually checks the real-time state of the slope dangerous rock through the VGA display screen.
As a further improvement of the invention, a complete moving target area is obtained based on a moving target detection algorithm of the self-adaptive mixed color space information fusion, full-automatic tracking is realized, the improvement of a color histogram template is completed, an S component is added into a color histogram only containing an H component, and an initial target two-dimensional color histogram template is established.
As a further improvement of the present invention, a moving object detection algorithm based on adaptive mixed color space information fusion includes:
Acquiring a video image sequence of real-time monitoring, converting video frame images from an RGB color space to HSV and Lab color spaces, and recombining different color channels to construct a mixed color space RS;
In a mixed color space RS, a self-adaptive multi-frame difference algorithm is utilized to realize the detection of a moving target with optimal performance, namely, a three-frame difference method is adopted to analyze a video image sequence in an initial frame, when the number of moving pixel points in the image is less than or equal to a research judgment threshold value, the interference of noise is judged, and the three-frame difference method is continuously used for detection at the moment; otherwise, judging that a large area of motion occurs in the scene, namely generating a moving object to be detected, and jumping to a five-frame difference method at the moment to continuously complete analysis of the video image sequence;
When the five-frame difference method is used for completing target detection, a Canny edge detection algorithm is utilized for extracting target edge information from binary images in the difference process;
fusing the five frames of differential information with the target edge information to increase the amount of fused information;
Extracting a target contour, performing research and judgment on the contour size, and deleting the contour when the contour size is undersized; and when the outline size meets the requirement, extracting the minimum circumscribed rectangular area of the outline from the original video image sequence, taking the area as a moving target area, storing the moving target area in a picture form for the use of a subsequent related algorithm, and immediately starting a moving target tracking algorithm.
As a further improvement of the invention, the Camshift algorithm is combined with the SURF feature point matching algorithm, when the situation that the target is re-appeared after being blocked by the obstacle occurs, the target repositioning is realized by utilizing the SURF feature point matching principle, the actual position area of the target is obtained and is transferred to the improved Camshift algorithm as a search window area, the video image sequence is analyzed, the position information of each tracking window is recorded in real time, and when a certain target is judged to stop moving or exceed the video image range, the target is considered to reach the tracking ending condition; when all targets reach the tracking end condition, the moving target tracking state should be ended and the moving target detection state should be skipped again.
As a further improvement of the present invention, a moving object tracking algorithm based on the improved continuous adaptive mean tracking cam shift algorithm includes: acquiring a real-time monitoring video image sequence, acquiring a complete moving target area according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, and storing the area in a picture form to serve as a template image of a SURF feature point matching algorithm, wherein the area is also used as an improved Camshift algorithm initial search window area;
converting the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculating an H-S two-dimensional color histogram in a search window;
Acquiring a color probability distribution map in a search window by utilizing the H-S two-dimensional color histogram, performing iterative computation by utilizing a Meanshift algorithm, and updating the position coordinates and the scale parameters of the search window when the convergence condition is met;
judging the position information of the search window, if the position variation of the continuous N frames of windows is smaller than epsilon, judging that the target stops moving and ending the tracking process of the target;
When the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful; when the size of the search window is not in the threshold value range, directly judging that the target is lost;
When the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended;
when the target is determined to be lost, the continuous video frame number of the lost target is studied and judged, if the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
When the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
As a further improvement of the present invention, the analysis and pre-judgment of the slope dangerous rock stability by the prediction model includes:
establishing a prediction model, carrying out result evaluation, comparing an output result with actual data, and simulating a result by the prediction model;
And (3) analyzing the result, determining whether the current state of the slope is normal or not by analyzing the reasoning result, and predicting and early warning the possible abnormal situation.
In order to achieve the above purpose, the present invention further provides the following technical solutions:
The utility model provides a flood discharge building exit area side slope dangerous rock monitoring system, its is applied to flood discharge building exit area side slope dangerous rock monitoring method, flood discharge building exit area side slope dangerous rock monitoring system includes:
The video image acquisition unit is used for acquiring a field video image sequence of the dangerous rock area in real time;
the video image processing unit is used for analyzing and judging video data through a moving target detection algorithm and a moving target tracking algorithm, acquiring image information of moving dangerous rock in time when the dangerous rock has an abnormal state, sending out an alarm and storing the processed field picture so as to be convenient for the call of related departments;
and the video image display unit is used for visually checking the real-time state of the slope dangerous rock through the VGA display screen.
As a further improvement of the present invention, a video image processing unit includes:
The image conversion subunit is used for acquiring a video image sequence monitored in real time, converting video frame images from RGB color space to HSV and Lab color space, and recombining different color channels to construct a proper mixed color space RS;
The image analysis subunit is used for realizing the detection of the moving target with optimal performance by utilizing a self-adaptive multi-frame difference algorithm in the mixed color space RS; the method comprises the steps that a video image sequence is analyzed by adopting a three-frame difference method in a starting frame, when the number of moving pixel points in the image is smaller than or equal to a research judgment threshold value, the image is judged to be interference of noise, and at the moment, the three-frame difference method is continuously used for detection; otherwise, judging that a large area of motion occurs in the scene, namely generating a moving object to be detected, and jumping to a five-frame difference method at the moment to continuously complete analysis of the video image sequence;
The information extraction subunit is used for extracting target edge information from the binary image in the differential process by utilizing a Canny edge detection algorithm when the target detection is completed by using a five-frame differential method;
the information fusion subunit is used for fusing the five-frame differential information with the target edge information to increase the fusion information quantity;
The contour extraction subunit is used for extracting a target contour, researching and judging the contour size, and deleting the contour when the contour size is undersized; and when the outline size meets the requirement, extracting the minimum circumscribed rectangular area of the outline from the original video image sequence, taking the area as a moving target area, storing the moving target area in a picture form for the use of a subsequent related algorithm, and immediately starting a moving target tracking algorithm.
As a further improvement of the present invention, the video image processing unit further includes:
The moving target region acquisition subunit is used for acquiring a real-time monitoring video image sequence, acquiring a complete moving target region according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, storing the region in a picture form to be used as a template image of a SURF feature point matching algorithm, and meanwhile, the region is also used as an improved Camshift algorithm initial search window region;
a color space conversion subunit, configured to convert the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculate an H-S two-dimensional color histogram in the search window;
The color probability distribution map acquisition subunit is used for acquiring a color probability distribution map in the search window by utilizing the H-S two-dimensional color histogram, performing iterative computation by utilizing a Meanshift algorithm, and updating the position coordinates and the scale parameters of the search window when the convergence condition is met;
The tracking process judging subunit is used for judging the position information of the search window, and judging that the target stops moving and finishes the tracking process of the target if the position variation of the continuous N frames of windows is smaller than epsilon;
When the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful;
when the size of the search window is not in the threshold value range, directly judging that the target is lost; when the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended;
When the target is determined to be lost, the number of frames of the video with the continuous target loss is studied and judged; if the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
When the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
According to the invention, the video monitoring instruments are arranged inside and outside the side slope dangerous rock, so that the monitoring data of the side slope dangerous rock can be acquired in real time, the monitoring image is generated, the state information of the side slope dangerous rock can be acquired in time, and the real-time monitoring of the side slope dangerous rock is realized; by adopting a moving target detection algorithm based on self-adaptive mixed color space information fusion, a moving target area of the slope dangerous rock can be automatically extracted, and compared with a traditional manual monitoring method, the automatic monitoring method has higher efficiency and accuracy; the moving target tracking algorithm based on the improved continuous self-adaptive mean tracking Camshift algorithm can track and position the dangerous rock of the side slope in real time and accurately, so that the position and the movement condition of the dangerous rock of the side slope can be accurately known, and a reliable data basis is provided for subsequent analysis and prediction.
According to the invention, through real-time monitoring, automatic detection, accurate tracking and stability analysis, the state of dangerous rock of the side slope can be comprehensively known, potential side slope disaster risks can be early warned in time, the safe operation of the flood discharge building is ensured, and scientific basis is provided for related decisions; the operation state and information of the side slope engineering can be displayed in real time, the scientificalness, the precision and the intellectualization level of the operation management of the hydraulic engineering are improved, and the operation and the maintenance of the hydraulic engineering are facilitated.
Drawings
FIG. 1 is a schematic flow chart of steps of an embodiment of a method for monitoring slope dangerous rock in an exit area of a flood discharge building according to the present invention;
FIG. 2 is a schematic diagram of a flow chart of monitoring images of slope dangerous rock generated by one embodiment of the method for monitoring slope dangerous rock in the exit area of the flood discharge building;
Fig. 3 is a schematic flow chart of a moving target detection algorithm based on adaptive mixed color space information fusion according to an embodiment of the method for monitoring slope dangerous rock in an exit area of a flood discharge building;
Fig. 4 is a schematic diagram of a moving object detection algorithm based on adaptive mixed color space information fusion according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart of a moving target tracking algorithm based on an improved continuous adaptive mean tracking Camshift algorithm according to one embodiment of the present invention;
FIG. 6 is a schematic diagram of a moving object tracking algorithm based on an improved continuous adaptive mean tracking Camshift algorithm according to one embodiment of the present invention for monitoring slope dangerous rock at the exit area of a flood discharge building;
FIG. 7 is a schematic flow chart of determining whether the current state of the slope dangerous rock is normal according to one embodiment of the monitoring method of the slope dangerous rock in the exit area of the flood discharge building;
fig. 8 is a schematic diagram of a flow chart of establishing a prediction model of an embodiment of a method for monitoring slope dangerous rock in an exit area of a flood discharge building according to the present invention;
fig. 9 is a schematic diagram of a flow chart of evaluation of a three-dimensional scene of side slope dangerous rock through a prediction model interface according to an embodiment of the method for monitoring side slope dangerous rock in an exit area of a flood discharge building;
fig. 10 is a schematic functional block diagram of an embodiment of a slope dangerous rock monitoring system in an exit area of a flood discharge building according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The terms "first," "second," "third," and the like in this disclosure are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "first," "second," and "third" may explicitly or implicitly include at least one such feature. In the description of the present invention, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise. All directional indications (such as up, down, left, right, front, back … …) in the embodiments of the present invention are merely used to explain the relative positional relationship, movement, etc. between the components in a particular gesture (as shown in the drawings), and if the particular gesture changes, the directional indication changes accordingly. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
As shown in fig. 1, the embodiment provides an embodiment of a method for monitoring slope dangerous rock in an exit area of a flood discharge building, which specifically includes the following steps:
Step S1: distributing video monitoring instruments in the inner part and the outer part of the side slope dangerous rock, collecting monitoring data of the side slope dangerous rock in real time, and generating a monitoring image of the side slope dangerous rock;
step S2: acquiring a real-time monitoring image, extracting a moving target area from the dangerous rock of the side slope based on a moving target detection algorithm fused by self-adaptive mixed color space information, and completing the real-time monitoring of the dangerous rock of the side slope;
step S3: analyzing each frame of real-time monitoring image of the side slope dangerous rock by taking the real-time monitoring image time factor of the side slope dangerous rock as input, and carrying out real-time and accurate tracking and positioning on the side slope dangerous rock based on a moving target tracking algorithm of an improved continuous self-adaptive mean tracking Camshift algorithm;
step S4: analyzing and pre-judging the stability of the side slope dangerous rock through a prediction model, determining whether the current state of the side slope dangerous rock is normal, and predicting and pre-warning;
Preferably, the steps S1 to S4 of the present embodiment implement real-time monitoring: by arranging the video monitoring instruments inside and outside the side slope dangerous rock, the monitoring data of the side slope dangerous rock can be acquired in real time, and the monitoring image is generated, so that the state information of the side slope dangerous rock can be acquired in time, and the real-time monitoring of the side slope dangerous rock is realized. And (3) automatic monitoring: by adopting a moving target detection algorithm based on self-adaptive mixed color space information fusion, a moving target region of slope dangerous rock can be automatically extracted, and compared with a traditional manual monitoring method, the automatic monitoring method has higher efficiency and accuracy. Accurate tracking and positioning: the moving target tracking algorithm based on the improved continuous self-adaptive mean tracking Camshift algorithm can track and position the dangerous rock of the side slope in real time and accurately, so that the position and the movement condition of the dangerous rock of the side slope can be accurately known, and a reliable data basis is provided for subsequent analysis and prediction. Stability analysis and prediction: the stability of the side slope dangerous rock is analyzed and prejudged through the prediction model, whether the current state of the side slope dangerous rock is normal can be judged, prediction and early warning are carried out, corresponding measures can be timely taken, the side slope disasters which can happen are prevented, and the safe operation of the flood discharge building is guaranteed.
In summary, the embodiment can comprehensively understand the state of dangerous rock of the side slope through real-time monitoring, automatic detection, accurate tracking and stability analysis, timely early warn potential side slope disaster risks, ensure safe operation of the flood discharge building and provide scientific basis for related decisions; the operation state and information of the side slope engineering can be displayed in real time, the scientificalness, the precision and the intellectualization level of the operation management of the hydraulic engineering are improved, and the operation and the maintenance of the hydraulic engineering are facilitated.
Further, as shown in fig. 2, the step S1 specifically includes the following steps:
s11: a hardware platform of the side slope dangerous rock monitoring system is built through the mutual matching of the video image acquisition unit, the video image processing unit and the video image display unit;
S12: a real-time video image sequence of the side slope dangerous rock area is acquired in real time through a video image acquisition unit; the video image processing unit analyzes and judges the video image through a moving target detection algorithm and a moving target tracking algorithm, when the abnormal state of the slope dangerous rock occurs, the video image processing unit timely acquires the image information of the slope dangerous rock movement, sends out an alarm and stores the processed field picture so as to be convenient for the call of related departments;
S13: the video image display unit visually checks the real-time state of the slope dangerous rock through the VGA display screen;
Preferably, the video image acquisition unit is composed of high-definition industrial cameras, and aiming at the characteristics of the monitoring environment of the dangerous rock of the side slope and the index parameters of various cameras, the system monitoring requirement is ensured.
The video image processing unit adopts a high-performance UP2 development board to process video data. Multiple codec modes are supported, and multiple operating systems are supported, including Windows 10, ubilinux, ubuntu, android, etc.
According to the embodiment, the real-time monitoring and early warning of the flood discharge building exit area side slope dangerous rock are realized by using a video monitoring technology and an image processing algorithm. The specific meaning includes: the safety is improved: collapse of dangerous rock on a side slope or landslide can cause plugging or damage to the exit area of a flood discharge building, resulting in casualties and property loss. Through real-time monitoring and early warning, measures can be taken in time, and the safety of personnel and the normal operation of the flood discharge building are ensured. Monitoring efficiency is improved: traditional slope dangerous rock monitoring methods generally require manual inspection, are time-consuming and not timely enough. By the video monitoring technology, all-weather and real-time monitoring of the side slope dangerous rock can be realized, and the monitoring efficiency and accuracy are greatly improved. The early warning capability is improved: the motion state of the side slope dangerous rock can be analyzed in real time through a moving target detection and tracking algorithm, and abnormal conditions are found and an alarm is sent out. This can help the relevant departments take timely measures to avoid potential disasters. The labor cost is reduced: the traditional slope dangerous rock monitoring method requires a great deal of manpower investment and has subjective factors. And through automatic video monitoring system, can reduce the manpower input to the risk of manual misjudgement has been reduced. Providing scientific basis: by collecting and analyzing the monitoring data of the side slope dangerous rock, a prediction model can be established, and the stability of the side slope dangerous rock is analyzed and predicted. The method provides scientific basis for related departments, and is helpful for making corresponding disaster prevention measures and emergency plans.
In summary, the meaning of the embodiment is that the monitoring and early warning capability of the slope dangerous rock of the exit area of the flood discharge building is improved, the safety and efficiency are improved, meanwhile, the labor cost is reduced, scientific basis is provided for related departments, and the normal operation of the flood discharge building and the safety of personnel are ensured.
Further, as shown in fig. 3, the step S2 specifically includes the following steps:
Step 21: acquiring a video image sequence monitored in real time, converting video frame images from an RGB (red green blue, which is a color model) color space to an HSV (hue, saturation and brightness, HSV is a color space model), lab (color space, brightness (L) and color components from red to green (a-axis) and from yellow to blue (b-axis)) color space, recombining different color channels to construct a mixed color space RS (describing the recombination of red, green and blue channels to create a new color space) suitable for the system;
Step 22: in the mixed color space RS, the motion target detection with optimal performance is realized by utilizing a self-adaptive multi-frame difference algorithm. The method comprises the steps that a video image sequence is analyzed by adopting a three-frame difference method in a starting frame, when the number of moving pixel points in the image is smaller than or equal to a research judgment threshold value, the image is judged to be interference of noise, and at the moment, the three-frame difference method is continuously used for detection; otherwise, judging that a large area of motion occurs in the scene (namely, judging that the number of motion pixel points in the scene is larger than the motion when the judging threshold value is exceeded), namely, generating a motion object to be detected, and then jumping to a five-frame difference method to continuously complete the analysis of the video image sequence;
Step 23: when the five-frame difference method is used for completing target detection, a Canny (Canny) edge detection algorithm is used for extracting target edge information of a binary image in the difference process, so that the integrity of a target contour is improved;
step 24: the five frames of differential information and the target edge information are fused to increase the amount of fusion information, so that the detection precision is improved;
step 25: extracting a target contour, studying and judging the contour size, deleting the contour when the contour size is undersized, and avoiding interference caused by external environmental factors or camera shake; and when the outline size meets the requirement, extracting the minimum circumscribed rectangular area of the outline from the original video image sequence, taking the area as a moving target area, storing the moving target area in a picture form for the use of a subsequent related algorithm, and immediately starting a moving target tracking algorithm.
Preferably, based on the principle of the moving object detection algorithm of adaptive mixed color space information fusion, referring to fig. 4, the design mainly has four parts: and acquiring a complete moving target area according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, replacing a process of manually confining the region of interest, reducing human errors and realizing the requirement of full-automatic tracking. The improvement of the color histogram template is completed. And adding an S component into the color histogram only containing the H component, establishing an initial target two-dimensional color histogram template, improving the utilization rate of color information, and improving the accuracy of color distinction.
Further, as shown in fig. 5, the step S3 specifically includes the following steps:
Step S31: acquiring a real-time monitoring video image sequence, acquiring a complete moving target area according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, and storing the area in a picture form to serve as a template image of a SURF (speedup robust feature) feature point matching algorithm, wherein the area is also taken as an initial search window area of an improved Camshift (continuous self-adaptive translation and scale estimation) algorithm;
Step 32: converting the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculating an H-S two-dimensional color histogram in a search window;
Step 33: acquiring a color probability distribution map in a search window by utilizing an H-S two-dimensional color histogram, carrying out iterative computation by utilizing a Meanshift (mean shift) algorithm, wherein the Meanshift algorithm is mainly used for image segmentation and target tracking, and updating the position coordinates and scale parameters of the search window when convergence conditions are met;
Step 34: judging the position information of the search window, if the position variation of the continuous N frames of windows is smaller than epsilon, judging that the target stops moving and ending the tracking process of the target;
Step 35: when the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful; when the size of the search window is not in the threshold value range, directly judging that the target is lost;
step 36: when the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended;
Step 37: when the target is determined to be lost, the number of frames of the video with the continuous target loss is studied and judged. If the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
Step 38: when the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
Preferably, the present embodiment proposes an improved Camshift algorithm for merging SURF feature point matching based on the principle of the moving object detection algorithm for adaptive mixed color space information fusion with reference to fig. 6. The Camshift algorithm is combined with the SURF feature point matching algorithm, when the situation that the target is re-appeared after being blocked by the obstacle occurs, the target repositioning is realized by utilizing the SURF feature point matching principle, the actual position area of the target is obtained and is transmitted to the improved Camshift algorithm to serve as a search window area, and the tracking capability of the algorithm under a complex scene is improved. And developing a closed-loop algorithm design of moving target detection and tracking. Analyzing the video image sequence, recording the position information of each tracking window in real time, and considering that a certain target reaches the tracking ending condition when judging that the target stops moving or exceeds the video image range; when all targets reach the tracking end condition, the moving target tracking state is required to be ended and the moving target detection state is required to be skipped again, so that the integrity of the software algorithm of the whole slope dangerous rock monitoring system is ensured.
Further, as shown in fig. 7, the step S4 specifically includes the following steps:
Step 41: establishing a prediction model, carrying out result evaluation, comparing an output result with actual data, and simulating the accuracy, stability and reliability of the result by the prediction model;
Step 42: and (3) analyzing the result, determining whether the current state of the slope is normal or not by analyzing the reasoning result, and predicting and early warning the possible abnormal situation.
Preferably, through the trained prediction model, under the operation of engineering personnel, the input parameters are used for obtaining a simulated digital twin model, so that the observation is convenient, and the engineering personnel can be helped to predict and simulate the data of the slope engineering under different working conditions.
Further, as shown in fig. 8, the establishment of the prediction model in step S41 includes the steps of:
step S410: acquiring video images of real-time monitoring of the side slope dangerous rock, acquiring a three-dimensional scene window corresponding to a video image analysis instruction, acquiring scene description parameters corresponding to the three-dimensional scene window, wherein the scene description parameters comprise the position and view angle parameters of the side slope dangerous rock, acquiring a scene description parameter time stamp, and storing the scene description parameters and the corresponding time stamp;
Step S411: generating a prediction model frame by codes according to a prediction model interface data structure and a prediction model interface definition, compiling based on the prediction model frame and a prediction program to generate a prediction model file, wherein the prediction model interface definition is described by adopting a text format of an XML language and comprises a prediction model name, a prediction model description, input interface information and output interface information; the prediction model framework comprises a definition common interface data structure and a calling interface function; calling an interface function for inputting input data in the form of parameters;
Step S412: and evaluating the three-dimensional scene of the side slope dangerous rock through a prediction model interface, simulating and analyzing the prediction model according to the video image and the three-dimensional scene of the side slope dangerous rock, evaluating whether the current state of the side slope dangerous rock is normal, and predicting and early warning possible abnormal conditions.
Preferably, in the embodiment, a video image monitored in real time is combined with a three-dimensional scene, and evaluation and prediction are performed through scene description parameters and a prediction model, so that state monitoring and early warning of slope dangerous rock are realized. Specifically, the meaning of the scheme includes: and (3) real-time monitoring: by acquiring the real-time monitoring video image of the side slope dangerous rock, the state change of the side slope dangerous rock can be known in time, and the potential dangerous situation can be found out early. Three-dimensional scene window: through the three-dimensional scene window corresponding to the video image analysis instruction, the state of the side slope dangerous rock can be observed and analyzed more intuitively, and more comprehensive information is provided. Scene description parameters: by collecting scene description parameters corresponding to the three-dimensional scene window, the position and the visual angle of the side slope dangerous rock can be accurately positioned, and accurate input data can be provided for subsequent evaluation and prediction. Prediction model framework: by generating the prediction model framework and the prediction model file, the evaluation and prediction can be performed based on the prediction model interface, and the accuracy and reliability of the dangerous rock state of the side slope are improved. Slope dangerous rock assessment and prediction: the three-dimensional scene of the side slope dangerous rock is evaluated through the prediction model interface, so that the state of the side slope dangerous rock can be simulated and analyzed, whether the side slope dangerous rock is normal or not is judged, and possible abnormal conditions are predicted and early-warned, so that a basis is provided for safety management and emergency treatment.
In summary, the significance of the embodiment is that the method combines real-time monitoring, three-dimensional scene and prediction model to provide comprehensive evaluation and prediction of dangerous rock state of side slope, thereby being beneficial to early finding potential dangerous situation and guaranteeing safety of personnel and property.
Further, as shown in fig. 9, the specific step S412 includes the following steps:
Step S4121: collecting video images and three-dimensional scene data of the side slope dangerous rock, wherein preprocessing such as image denoising, scene point cloud processing and the like is needed; extracting target features from the acquired data; the video image is used for extracting the shape, color, texture and other characteristics of the slope dangerous rock by using a computer vision technology; for a three-dimensional scene, extracting the characteristics of the height, gradient, curvature and the like of a side slope;
Step S4122: labeling the extracted target features, namely labeling each video image and three-dimensional scene data with normal or abnormal labels; then training a prediction model by using the marked data set; testing and evaluating part of the data by using the trained model to evaluate the accuracy and performance of the model;
Step S4123: simulating and analyzing the video image and the three-dimensional scene of the new slope dangerous rock by using the trained prediction model; the prediction model predicts whether the state of the current slope dangerous rock is normal or not and the possible abnormal situation according to the extracted characteristics and the previous training result; the results of the simulation and analysis may be two categories (normal/abnormal) or multiple categories (normal/mild abnormal/severe abnormal).
Preferably, through the steps, the prediction model can be used for monitoring and predicting the side slope dangerous rock in real time, abnormal conditions can be found timely, corresponding measures can be taken, and the safety and stability of the side slope dangerous rock are improved. The three-dimensional scene of the side slope dangerous rock is evaluated through the prediction model interface, so that the prediction and analysis of the side slope dangerous rock state are realized, the method has important significance for monitoring and early warning of the side slope dangerous rock, and people can be helped to take measures in time to avoid or reduce dangerous accidents; by collecting video images and three-dimensional scene data of the side slope dangerous rock and extracting features, a prediction model can be trained to judge whether the state of the side slope dangerous rock is normal or not, and meanwhile, possible abnormal conditions can be simulated and analyzed; the abnormal condition of dangerous rock of the side slope can be found in advance, preventive measures can be taken in time, and the life and property safety of people is guaranteed.
As shown in fig. 10, this embodiment also provides an embodiment of a flood discharge building exit area slope dangerous rock monitoring system, in this embodiment, the flood discharge building exit area slope dangerous rock monitoring system is applied to the method of the flood discharge building exit area slope dangerous rock monitoring system in the above embodiment, and the flood discharge building exit area slope dangerous rock monitoring system includes a video image acquisition unit 1, a video image processing unit 2 and a video image display unit 3 electrically connected in sequence.
The video image acquisition unit 1 acquires a field video image sequence of a dangerous rock area in real time; the video image processing unit 2 analyzes and judges the video data through a moving target detection algorithm and a moving target tracking algorithm, when the dangerous rock mass has an abnormal state, the video image processing unit timely acquires the image information of the moving dangerous rock, sends out an alarm and stores the processed field pictures so as to be convenient for the call of related departments; the video image display unit 3 visually checks the real-time state of the slope dangerous rock through the VGA display screen.
Further, the video image processing unit comprises an image conversion subunit, an image analysis subunit, an information extraction subunit, an information fusion subunit and a contour extraction subunit which are electrically connected in sequence.
Wherein the image conversion subunit is configured to obtain a video image sequence monitored in real time, convert the video frame image from RGB (red, green and blue, which is a color model) color space to HSV (hue, saturation and brightness, HSV is a color space model), lab (color space, brightness (L) and color components from red to green (a-axis) and from yellow to blue (b-axis)) color space, and recombine different color channels to construct a mixed color space RS (describing recombination of red, green and blue channels to create a new color space) suitable for the present system;
The image analysis subunit is used for realizing the moving target detection with optimal performance by utilizing the self-adaptive multi-frame difference algorithm in the mixed color space RS. The method comprises the steps that a video image sequence is analyzed by adopting a three-frame difference method in a starting frame, when the number of moving pixel points in the image is smaller than or equal to a research judgment threshold value, the image is judged to be interference of noise, and at the moment, the three-frame difference method is continuously used for detection; otherwise, judging that a large area of motion occurs in the scene, namely generating a moving object to be detected, and jumping to a five-frame difference method at the moment to continuously complete analysis of the video image sequence;
The information extraction subunit is used for extracting target edge information of the binary image in the differential process by utilizing a Canny edge detection algorithm when the target detection is completed by using a five-frame differential method, so that the integrity of the target contour is improved;
the information fusion subunit is used for fusing the five-frame differential information with the target edge information to increase the amount of fusion information and improve the detection precision;
The contour extraction subunit is used for extracting a target contour, researching and judging the contour size, deleting the contour when the contour size is undersized, and avoiding interference caused by external environment factors or camera shake; and when the outline size meets the requirement, extracting the minimum circumscribed rectangular area of the outline from the original video image sequence, taking the area as a moving target area, storing the moving target area in a picture form for the use of a subsequent related algorithm, and immediately starting a moving target tracking algorithm.
Further, the video image processing unit further comprises a moving target region acquisition subunit, a color space conversion subunit, a color probability distribution map acquisition subunit and a tracking process judgment subunit which are electrically connected in sequence.
The moving target region acquisition subunit is used for acquiring a real-time monitoring video image sequence, acquiring a complete moving target region according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, storing the region in a picture form to be used as a template image of a SURF (speedup robust feature) feature point matching algorithm, and meanwhile, the region is also used as an initial search window region of an improved Camshift (continuous self-adaptive translation and scale estimation) algorithm;
The color space conversion subunit is used for converting the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculating an H-S two-dimensional color histogram in the search window;
The color probability distribution map acquisition subunit is used for acquiring a color probability distribution map in the search window by utilizing the H-S two-dimensional color histogram, and updating the position coordinates and the scale parameters of the search window when the convergence condition is met by utilizing the Meanshift (mean shift, mainly used for image segmentation and target tracking) algorithm iterative computation;
The tracking process judging subunit is used for judging the position information of the search window, and judging that the target stops moving and finishes the tracking process of the target if the position variation of the continuous N frames of windows is smaller than epsilon; when the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful; when the size of the search window is not in the threshold value range, directly judging that the target is lost; when the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended; when the target is determined to be lost, the number of frames of the video with the continuous target loss is studied and judged. If the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
When the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
Further, the video image processing unit further comprises a result evaluation subunit and a result output subunit which are electrically connected in sequence.
The result evaluation subunit is used for establishing a prediction model, evaluating the result, comparing the output result with actual data, and simulating the accuracy, stability and reliability of the result by the prediction model;
The result output subunit is used for analyzing the result, determining whether the current state of the slope is normal or not through analyzing and reasoning the result, and predicting and early warning the possible abnormal situation.
The embodiment can monitor the state of the dangerous rock of the side slope in real time by utilizing a video image acquisition and processing technology, and give an alarm in time when an abnormal state occurs, so as to provide image information called by related departments. The system method can effectively help monitoring staff to know the real-time state of the dangerous rock of the side slope, predicts and early warns possible abnormal conditions, is favorable for taking timely measures to treat and prevent, and ensures the safety of the side slope.
In the several embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of elements is merely a logical functional division, and there may be additional divisions of actual implementation, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other forms.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units. The foregoing is only the embodiments of the present invention, and the patent scope of the invention is not limited thereto, but is also covered by the patent protection scope of the invention, as long as the equivalent structure or equivalent flow changes made by the description and the drawings of the invention or the direct or indirect application in other related technical fields are adopted.
The embodiments of the invention have been described in detail above, but they are merely examples, and the invention is not limited to the above-described embodiments. It will be apparent to those skilled in the art that any equivalent modifications or substitutions to this invention are within the scope of the invention, and therefore, all equivalent changes and modifications, improvements, etc. that do not depart from the spirit and scope of the principles of the invention are intended to be covered by this invention.
Claims (3)
1. The monitoring method for the slope dangerous rock in the outlet area of the flood discharge building is characterized by comprising the following steps of:
distributing video monitoring instruments in the inner part and the outer part of the side slope dangerous rock, collecting monitoring data of the side slope dangerous rock in real time, and generating a monitoring image of the side slope dangerous rock;
Acquiring a real-time monitoring image, extracting a moving target area from the dangerous rock of the side slope based on a moving target detection algorithm fused by self-adaptive mixed color space information, and completing the real-time monitoring of the dangerous rock of the side slope;
Analyzing each frame of real-time monitoring image of the side slope dangerous rock by taking the real-time monitoring image time factor of the side slope dangerous rock as input, and carrying out real-time and accurate tracking and positioning on the side slope dangerous rock based on a moving target tracking algorithm of an improved continuous self-adaptive mean tracking Camshift algorithm;
Analyzing and pre-judging the stability of the side slope dangerous rock through a prediction model, determining whether the current state of the side slope dangerous rock is normal, and predicting and pre-warning;
Combining a Camshift algorithm with a SURF feature point matching algorithm, realizing target repositioning by utilizing the SURF feature point matching principle when a situation occurs that a target is re-appeared after being blocked by an obstacle, obtaining an actual position area of the target, transmitting the actual position area to an improved Camshift algorithm as a search window area, analyzing a video image sequence, recording the position information of each tracking window in real time, and considering that the target reaches a tracking end condition when judging that a certain target stops moving or exceeds the video image range; when all targets reach the tracking end condition, the tracking state of the moving target is required to be ended and the moving target detection state is required to be skipped again;
Analyzing and pre-judging the stability of the side slope dangerous rock through a prediction model, wherein the method comprises the following steps of:
establishing a prediction model, carrying out result evaluation, comparing an output result with actual data, and simulating a result by the prediction model;
analyzing results, namely determining whether the current state of the slope is normal or not through analyzing reasoning results, and predicting and early warning possible abnormal conditions;
the establishment of the prediction model comprises the following steps:
Acquiring video images of real-time monitoring of the side slope dangerous rock, acquiring a three-dimensional scene window corresponding to a video image analysis instruction, acquiring scene description parameters corresponding to the three-dimensional scene window, wherein the scene description parameters comprise the position and view angle parameters of the side slope dangerous rock, acquiring a scene description parameter time stamp, and storing the scene description parameters and the corresponding time stamp;
Generating a prediction model frame by codes according to a prediction model interface data structure and a prediction model interface definition, compiling based on the prediction model frame and a prediction program to generate a prediction model file, wherein the prediction model interface definition is described by adopting a text format of an XML language and comprises a prediction model name, a prediction model description, input interface information and output interface information; the prediction model framework comprises a definition common interface data structure and a calling interface function; calling an interface function for inputting input data in the form of parameters;
The three-dimensional scene of the side slope dangerous rock is evaluated through a prediction model interface, the prediction model is used for simulating and analyzing according to the video image and the three-dimensional scene of the side slope dangerous rock, whether the current state of the side slope dangerous rock is normal or not is evaluated, and possible abnormal conditions are predicted and early-warned; the method specifically comprises the following steps:
Collecting video images and three-dimensional scene data of the side slope dangerous rock, wherein preprocessing such as image denoising and scene point cloud processing is required; extracting target features from the acquired data; the method comprises the steps of (1) extracting shape, color and texture characteristics of slope dangerous rock from video images by using a computer vision technology; for a three-dimensional scene, extracting the height, gradient and curvature characteristics of a side slope;
Labeling the extracted target features, namely labeling each video image and three-dimensional scene data with normal or abnormal labels; then training a prediction model by using the marked data set; testing and evaluating part of the data by using the trained model to evaluate the accuracy and performance of the model;
simulating and analyzing the video image and the three-dimensional scene of the new slope dangerous rock by using the trained prediction model; the prediction model predicts whether the state of the current slope dangerous rock is normal or not and the possible abnormal situation according to the extracted characteristics and the previous training result; the result of the simulation and analysis is a two-class or multi-class;
Based on a moving target detection algorithm of self-adaptive mixed color space information fusion, a complete moving target area is obtained, full-automatic tracking is realized, improvement of a color histogram template is completed, an S component is added into a color histogram only containing an H component, and an initial target two-dimensional color histogram template is established;
The moving target detection algorithm based on the self-adaptive mixed color space information fusion comprises the following steps:
Acquiring a video image sequence of real-time monitoring, converting video frame images from an RGB color space to HSV and Lab color spaces, and recombining different color channels to construct a mixed color space RS;
In a mixed color space RS, a self-adaptive multi-frame difference algorithm is utilized to realize the detection of a moving target with optimal performance, namely, a three-frame difference method is adopted to analyze a video image sequence in an initial frame, when the number of moving pixel points in the image is less than or equal to a research judgment threshold value, the interference of noise is judged, and the three-frame difference method is continuously used for detection at the moment; otherwise, judging that the number of the motion pixel points in the scene is greater than the judging threshold value, namely generating a motion object to be detected, and jumping to a five-frame difference method at the moment to continuously finish analysis of the video image sequence;
When the five-frame difference method is used for completing target detection, a Canny edge detection algorithm is utilized for extracting target edge information from binary images in the difference process;
fusing the five frames of differential information with the target edge information to increase the amount of fused information;
Extracting a target contour, performing research and judgment on the contour size, and deleting the contour when the contour size is undersized; when the outline size meets the requirement, extracting a minimum circumscribed rectangular area of the outline from the original video image sequence, taking the minimum circumscribed rectangular area as a moving target area, storing the moving target area in a picture form, and then starting a moving target tracking algorithm;
a moving object tracking algorithm based on an improved continuous adaptive mean tracking cam shift algorithm, comprising: acquiring a real-time monitoring video image sequence, acquiring a complete moving target area according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, and storing the moving target area in a picture form to serve as a template image of a SURF feature point matching algorithm, wherein the moving target area is also taken as an improved Camshift algorithm initial search window area;
converting the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculating an H-S two-dimensional color histogram in a search window;
Acquiring a color probability distribution map in a search window by utilizing the H-S two-dimensional color histogram, performing iterative computation by utilizing a Meanshift algorithm, and updating the position coordinates and the scale parameters of the search window when the convergence condition is met;
judging the position information of the search window, if the position variation of the continuous N frames of windows is smaller than epsilon, judging that the target stops moving and ending the tracking process of the target;
When the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful; when the size of the search window is not in the threshold value range, directly judging that the target is lost;
When the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended;
when the target is determined to be lost, the continuous video frame number of the lost target is studied and judged, if the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
When the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
2. The method for monitoring slope dangerous rock in an exit area of a flood discharge building according to claim 1, wherein the step of collecting monitoring data of the slope dangerous rock in real time and generating a monitoring image of the slope dangerous rock comprises the steps of:
A hardware platform of the side slope dangerous rock monitoring system is built through the mutual matching of the video image acquisition unit, the video image processing unit and the video image display unit;
A real-time video image sequence of the side slope dangerous rock area is acquired in real time through a video image acquisition unit; the video image processing unit analyzes and judges the video image through a moving target detection algorithm and a moving target tracking algorithm, and when the abnormal state of the slope dangerous rock occurs, the video image processing unit timely acquires the image information of the slope dangerous rock movement, sends out an alarm and stores the processed field picture;
The video image display unit visually checks the real-time state of the slope dangerous rock through the VGA display screen.
3. A flood discharge building exit area slope dangerous rock monitoring system applied to the flood discharge building exit area slope dangerous rock monitoring method according to any one of claims 1 to 2, wherein the flood discharge building exit area slope dangerous rock monitoring system comprises:
The video image acquisition unit is used for acquiring a field video image sequence of the dangerous rock area in real time;
The video image processing unit is used for analyzing and judging video data through a moving target detection algorithm and a moving target tracking algorithm, timely acquiring image information of moving dangerous rock when the dangerous rock has an abnormal state, sending out an alarm and storing the processed field picture for calling;
the video image display unit is used for visually checking the real-time state of the slope dangerous rock through the VGA display screen;
The video image processing unit further comprises a result evaluation subunit and a result output subunit which are electrically connected in sequence;
the result evaluation subunit is used for establishing a prediction model, evaluating the result, comparing the output result with actual data, and simulating the accuracy, stability and reliability of the result by the prediction model;
The result output subunit is used for analyzing the result, determining whether the current state of the slope is normal or not through analyzing and reasoning the result, and predicting and early warning the possible abnormal situation;
The establishment of the prediction model comprises the following steps:
Acquiring video images of real-time monitoring of the side slope dangerous rock, acquiring a three-dimensional scene window corresponding to a video image analysis instruction, acquiring scene description parameters corresponding to the three-dimensional scene window, wherein the scene description parameters comprise the position and view angle parameters of the side slope dangerous rock, acquiring a scene description parameter time stamp, and storing the scene description parameters and the corresponding time stamp;
Generating a prediction model frame by codes according to a prediction model interface data structure and a prediction model interface definition, compiling based on the prediction model frame and a prediction program to generate a prediction model file, wherein the prediction model interface definition is described by adopting a text format of an XML language and comprises a prediction model name, a prediction model description, input interface information and output interface information; the prediction model framework comprises a definition common interface data structure and a calling interface function; calling an interface function for inputting input data in the form of parameters;
The three-dimensional scene of the side slope dangerous rock is evaluated through a prediction model interface, the prediction model is used for simulating and analyzing according to the video image and the three-dimensional scene of the side slope dangerous rock, whether the current state of the side slope dangerous rock is normal or not is evaluated, and possible abnormal conditions are predicted and early-warned; the method specifically comprises the following steps:
Collecting video images and three-dimensional scene data of the side slope dangerous rock, wherein preprocessing such as image denoising and scene point cloud processing is required; extracting target features from the acquired data; the method comprises the steps of (1) extracting shape, color and texture characteristics of slope dangerous rock from video images by using a computer vision technology; for a three-dimensional scene, extracting the height, gradient and curvature characteristics of a side slope;
Labeling the extracted target features, namely labeling each video image and three-dimensional scene data with normal or abnormal labels; then training a prediction model by using the marked data set; testing and evaluating part of the data by using the trained model to evaluate the accuracy and performance of the model;
simulating and analyzing the video image and the three-dimensional scene of the new slope dangerous rock by using the trained prediction model; the prediction model predicts whether the state of the current slope dangerous rock is normal or not and the possible abnormal situation according to the extracted characteristics and the previous training result; the result of the simulation and analysis is a two-class or multi-class;
a video image processing unit comprising:
The image conversion subunit is used for acquiring a video image sequence monitored in real time, converting video frame images from RGB color space to HSV and Lab color space, and recombining different color channels to construct a proper mixed color space RS;
the image analysis subunit is used for realizing the detection of the moving target with optimal performance by utilizing a self-adaptive multi-frame difference algorithm in the mixed color space RS; the method comprises the steps that a video image sequence is analyzed by adopting a three-frame difference method in a starting frame, when the number of moving pixel points in the image is smaller than or equal to a research judgment threshold value, the image is judged to be interference of noise, and at the moment, the three-frame difference method is continuously used for detection; otherwise, judging that the number of the motion pixel points in the scene is greater than the judging threshold value, namely generating a motion object to be detected, and jumping to a five-frame difference method at the moment to continuously finish analysis of the video image sequence;
The information extraction subunit is used for extracting target edge information from the binary image in the differential process by utilizing a Canny edge detection algorithm when the target detection is completed by using a five-frame differential method;
the information fusion subunit is used for fusing the five-frame differential information with the target edge information to increase the fusion information quantity;
The contour extraction subunit is used for extracting a target contour, researching and judging the contour size, and deleting the contour when the contour size is undersized; when the outline size meets the requirement, extracting a minimum circumscribed rectangular area of the outline from the original video image sequence, taking the minimum circumscribed rectangular area as a moving target area, storing the minimum circumscribed rectangular area in a picture form, and immediately starting a moving target tracking algorithm;
the video image processing unit further includes:
The moving target region acquisition subunit is used for acquiring a real-time monitoring video image sequence, acquiring a complete moving target region according to a moving target detection algorithm based on self-adaptive mixed color space information fusion, storing the region in a picture form to be used as a template image of a SURF feature point matching algorithm, and meanwhile, the region is also used as an improved Camshift algorithm initial search window region;
a color space conversion subunit, configured to convert the video frame image from RGB color space to HSV color space through a color space conversion formula, and calculate an H-S two-dimensional color histogram in the search window;
The color probability distribution map acquisition subunit is used for acquiring a color probability distribution map in the search window by utilizing the H-S two-dimensional color histogram, performing iterative computation by utilizing a Meanshift algorithm, and updating the position coordinates and the scale parameters of the search window when the convergence condition is met;
The tracking process judging subunit is used for judging the position information of the search window, and judging that the target stops moving and finishes the tracking process of the target if the position variation of the continuous N frames of windows is smaller than epsilon;
When the size of the search window area is within the set threshold value range, continuously calculating the Pasteur distance between the target area and the search window area, if the value is greater than the threshold value Tos, considering that the target is lost and tracked to fail, otherwise, considering that the target tracking is successful;
when the size of the search window is not in the threshold value range, directly judging that the target is lost; when the target tracking is judged to be successful, directly reading a new frame of video image sequence to continue tracking until the video image sequence is ended;
When the target is determined to be lost, the number of frames of the video with the continuous target loss is studied and judged; if the condition that the continuous M frames of targets are lost exists, the targets are considered to exceed the video image range, and the tracking process of the targets is ended; otherwise, performing target repositioning according to the SURF feature point matching algorithm, and then continuing to read a new frame of video image sequence for tracking until the video image sequence is ended;
When the moving targets meet the tracking end condition and the tracking process is finished, judging all the moving target states in the video image sequence, if all the targets meet the tracking end condition, stopping the moving target tracking state from jumping to the moving target detection state, otherwise, obtaining a new frame of video image sequence to continue tracking the targets which do not meet the tracking end condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410356023.XA CN117953016B (en) | 2024-03-27 | 2024-03-27 | Flood discharge building exit area slope dangerous rock monitoring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410356023.XA CN117953016B (en) | 2024-03-27 | 2024-03-27 | Flood discharge building exit area slope dangerous rock monitoring method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117953016A CN117953016A (en) | 2024-04-30 |
| CN117953016B true CN117953016B (en) | 2024-07-16 |
Family
ID=90792607
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410356023.XA Active CN117953016B (en) | 2024-03-27 | 2024-03-27 | Flood discharge building exit area slope dangerous rock monitoring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117953016B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118485959A (en) * | 2024-05-07 | 2024-08-13 | 浙江高信技术股份有限公司 | Data processing system for monitoring user safety state in building area |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112232138A (en) * | 2020-09-25 | 2021-01-15 | 天津大学 | Channel slope damage intelligent identification method based on superpixel characteristics |
| CN116385906A (en) * | 2023-03-10 | 2023-07-04 | 西南交通大学 | Side slope disaster light sensing system, equipment and platform based on deep learning |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5118590B2 (en) * | 2008-09-16 | 2013-01-16 | キヤノン株式会社 | Subject tracking method and imaging apparatus |
| CN107154053A (en) * | 2017-05-11 | 2017-09-12 | 南宁市正祥科技有限公司 | Moving target detecting method under static background |
| CN108020263A (en) * | 2017-11-21 | 2018-05-11 | 桂林理工大学 | A kind of precarious rock mass monitoring method based on wireless signal |
| CN110288538A (en) * | 2019-05-23 | 2019-09-27 | 南京理工大学 | A kind of the moving target shadow Detection and removing method of multiple features fusion |
| US20210312236A1 (en) * | 2020-03-30 | 2021-10-07 | Cherry Labs, Inc. | System and method for efficient machine learning model training |
| CN112818753B (en) * | 2021-01-11 | 2024-06-28 | 精英数智科技股份有限公司 | Method, device and system for detecting mining pit falling objects |
| CN114353876A (en) * | 2022-01-07 | 2022-04-15 | 兰州大学 | Loess highway slope health monitoring method |
| CN114639041A (en) * | 2022-03-16 | 2022-06-17 | 南京航空航天大学 | Flame detection method based on motion characteristics and color characteristics |
| CN116109701A (en) * | 2022-12-23 | 2023-05-12 | 洛阳中科人工智能研究院有限公司 | Object grabbing method based on passive dual-purpose high light reflection |
| CN116229248A (en) * | 2023-03-10 | 2023-06-06 | 中科曙光国际信息产业有限公司 | Ocean species distribution prediction method, device, equipment and storage medium |
| CN117034789A (en) * | 2023-05-18 | 2023-11-10 | 河北建筑工程学院 | Typical landslide prediction method under rainwater infiltration effect |
| CN117058845A (en) * | 2023-08-07 | 2023-11-14 | 河南交院工程技术集团有限公司 | Highway high slope landslide hazard early warning method based on evaluation function training |
| CN117671508B (en) * | 2024-02-01 | 2024-04-05 | 中国水利水电第七工程局有限公司 | SAR image-based high-steep side slope landslide detection method and system |
| CN117765708A (en) * | 2024-02-22 | 2024-03-26 | 城云科技（中国）有限公司 | Slope instability prediction method and device based on map image analysis technology |
-
2024
- 2024-03-27 CN CN202410356023.XA patent/CN117953016B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112232138A (en) * | 2020-09-25 | 2021-01-15 | 天津大学 | Channel slope damage intelligent identification method based on superpixel characteristics |
| CN116385906A (en) * | 2023-03-10 | 2023-07-04 | 西南交通大学 | Side slope disaster light sensing system, equipment and platform based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| 基于目标检测与跟踪算法的边坡危岩监测系统设计;王扬;《硕士电子期刊》;20210115;第2-4节 * |
| 王扬.基于目标检测与跟踪算法的边坡危岩监测系统设计.《硕士电子期刊》.2021,第2-4节. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117953016A (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11634987B2 (en) | Safety early warning method and device for full-section tunneling of tunnel featuring dynamic water and weak surrounding rock | |
| CN110502965B (en) | Construction safety helmet wearing monitoring method based on computer vision human body posture estimation | |
| CN110602449A (en) | Intelligent construction safety monitoring system method in large scene based on vision | |
| CN103700261A (en) | Video-based road traffic flow feature parameter monitoring and traffic comprehensive information service system | |
| CN117953016B (en) | Flood discharge building exit area slope dangerous rock monitoring method and system | |
| CN111178424A (en) | Petrochemical production site safety compliance real-time detection system and method | |
| CN114812403A (en) | Large-span steel structure hoisting deformation monitoring method based on unmanned aerial vehicle and machine vision | |
| CN115326026B (en) | Method and device for acquiring hydraulic characteristics based on non-contact measurement-hydrodynamic fusion assimilation | |
| CN111896549A (en) | Building crack monitoring system and method based on machine learning | |
| CN104570764A (en) | Verification platform for airborne target indicating system | |
| CN113011280A (en) | Method and device for detecting person contact distance, computer equipment and storage medium | |
| CN112685812A (en) | Dynamic supervision method, device, equipment and storage medium | |
| KR102669058B1 (en) | SAFETY INSPECTION SYSTEM FOR STRUCTURE USING SMART PORTABLE DEVICE, AND METHOD FOR THe SAME | |
| CN116524017B (en) | Underground detection, identification and positioning system for mine | |
| CN116012336B (en) | Tunnel intelligent geological sketch and surrounding rock grade identification device and method | |
| CN113935859A (en) | Intelligent construction site system based on BIM model | |
| CN115035251B (en) | Bridge deck vehicle real-time tracking method based on field enhanced synthetic data set | |
| Zeng et al. | Vision-based high-precision intelligent monitoring for shield tail clearance | |
| CN118351654A (en) | Intelligent early warning method and system for geological disaster and engineering safety monitoring | |
| CN114495421A (en) | Intelligent open type road construction operation monitoring and early warning method and system | |
| Siddiqui | UWB RTLS for construction equipment localization: experimental performance analysis and fusion with video data | |
| CN116740610A (en) | AI algorithm system for highway construction safety management | |
| KR20220120211A (en) | Drone image analysis system based on deep learning for traffic measurement | |
| CN115409691A (en) | Bimodal learning slope risk detection method integrating laser ranging and monitoring image | |
| Zhou et al. | Construction Activity Analysis of Workers Based on Human Posture Estimation Information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |