CN113780113A - Pipeline violation behavior identification method - Google Patents
Pipeline violation behavior identification method Download PDFInfo
- Publication number
- CN113780113A CN113780113A CN202110983603.8A CN202110983603A CN113780113A CN 113780113 A CN113780113 A CN 113780113A CN 202110983603 A CN202110983603 A CN 202110983603A CN 113780113 A CN113780113 A CN 113780113A
- Authority
- CN
- China
- Prior art keywords
- pipeline
- aerial vehicle
- unmanned aerial
- identifying
- image
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 230000006399 behavior Effects 0.000 claims abstract description 17
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims description 13
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 230000002457 bidirectional effect Effects 0.000 claims description 6
- 230000000306 recurrent effect Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 5
- 230000004927 fusion Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000007499 fusion processing Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000012850 discrimination method Methods 0.000 claims 1
- 238000010276 construction Methods 0.000 description 8
- 230000000903 blocking effect Effects 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for identifying pipeline violation behaviors, which comprises the following steps of: s1, pre-flying of the unmanned aerial vehicle; collecting pipeline data in the process of pre-flying of the unmanned aerial vehicle, preprocessing images, filtering data which is not available on the periphery of a pipeline, identifying pipeline interface data, identifying whether workers have safety helmets, detecting the condition of people standing on the pipeline, and extracting key areas in pipeline detection; s2, controlling the unmanned aerial vehicle to fly to a key area, identifying a pipeline port, shooting the condition of a pipeline port plug in an oriented mode, and judging the plugging condition of the pipeline. The cloud deck is arranged in the pipeline violation behavior identification system, the plugging conditions of two ends of the pipeline can be detected simultaneously through the cloud deck in the flight process of the unmanned aerial vehicle, and the problem that the pipeline plug of the unmanned aerial vehicle in the navigation direction can only be detected in the flight process of the unmanned aerial vehicle in the prior art is solved.
Description
Technical Field
The invention relates to the technical field of pipeline detection, in particular to a pipeline violation behavior identification method.
Background
The oil and gas pipeline is used as an important tool for energy transportation, once leakage or explosion accidents occur, the safety of lives and properties of people is seriously threatened, and therefore the safety patrol and management of the pipeline are very important. The national oil and gas pipeline protection act stipulates: in the area ranges of five meters on two sides of the central line of the pipeline, the behaviors of planting deep-rooted plants, excavating construction, illegal construction and the like are forbidden. However, the above-mentioned actions that endanger the safety of oil and gas pipelines occur occasionally, even causing irreparable losses.
The existing pipeline supervision process is mainly used for supervising whether an oil-gas pipeline is blocked or not in the construction process and whether a person stands or not in the pipeline construction process. The pipeline construction method is mainly realized by manual monitoring at present, and the instantaneity is not strong. Especially, in the process of pipeline construction, whether a safety cap is arranged or not, whether plugs at two ends of the pipeline are plugged or not and the like exist or not, and the pipeline can be timely reminded after the problems are found. In the prior art, supervision is mainly performed in a manual mode, and the supervision has large workload and low real-time performance. Some units adopt a rotor type unmanned aerial vehicle to supervise the behavior, acquire data through a camera, and process various violation behaviors through a computer workstation by adopting an image processing method, so that on one hand, the real-time performance is not strong, and on the other hand, due to factors such as large data volume to be processed, the processing speed of computer hardware is high, and the cost is increased; and only can detect the pipeline end cap of unmanned aerial vehicle navigation direction at unmanned aerial vehicle flight in-process, can't detect the shutoff condition at pipeline both ends simultaneously.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a pipeline violation behavior identification method, which aims to solve the problems that a large amount of manpower is wasted and the instantaneity is not strong because the supervision in the pipeline construction process is mainly carried out manually, and the blocking conditions of two ends of a pipeline cannot be detected simultaneously in the flight process of an unmanned aerial vehicle.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows.
The pipeline violation behavior identification method comprises the following steps:
s1, pre-flying of the unmanned aerial vehicle; collecting pipeline data in the process of pre-flying of the unmanned aerial vehicle, preprocessing images, filtering data which is not available on the periphery of a pipeline, identifying pipeline interface data, identifying whether workers have safety helmets, detecting the condition of people standing on the pipeline, and extracting key areas in pipeline detection;
s2, controlling the unmanned aerial vehicle to fly to a key area, identifying a pipeline port, shooting the condition of a pipeline port plug in an oriented mode, and judging the plugging condition of the pipeline.
Further optimizing the technical scheme, and detecting the condition of the people on the pipeline by a pipeline people on station distinguishing method; the method for distinguishing the people standing on the pipeline comprises the following steps:
s101, collecting images of a plurality of frames of video without foreign matters within a period of time, summing the images and averaging the images to obtain an initial background image, namely the initial background image
Wherein, BI(x, y) is the initial background image, Fi(x, y) represents the ith frame image;
s102, obtaining a difference image by subtracting the current frame image from the background image, carrying out binarization processing on the difference image, and preliminarily separating a target from the background;
s103, filtering isolated foreground points in the binary image, performing expansion operation to obtain a complete target profile, and determining whether foreign matters exist on the pipeline or whether dangerous behaviors of people standing on the pipeline exist.
The technical scheme is further optimized, personnel come is judged through a personnel identification method, binarization processing is carried out, and whether the personnel are provided with safety helmets or not is judged through the top color of the heads.
Further optimizing the technical scheme, the method for identifying whether the worker carries the safety helmet comprises the following steps:
s111, constructing an MsxModule structure by using a YOLOv4-tiny lightweight neural network, and adding the MsxModule structure in a multi-scale fusion process to extract main characteristics of small and medium-sized targets;
s112, acquiring more main features by utilizing a Max Module structure, reserving edge information and positioning information of the shallow layer network by utilizing a bottom-up feature fusion structure, wherein a feedforward transfer equation is as follows:
the weight update equation is:
and (3) positions of the CIoU prediction frame and the real frame are quoted, so that the loss convergence speed is accelerated, and the loss function is as follows:
s113, constructing a data set of the wearable safety helmet, reconstructing the proportion of the anchor frames, and processing a training set.
Further optimizing the technical solution, the step S2 includes the following steps:
s201, directly positioning a pipeline port;
s202, effectively correcting the positioned inclined port;
s203, improving the CNN network;
s204, extracting time sequence characteristics by using the improved CNN network and the corrected picture;
s205, identifying the time sequence label by adopting a bidirectional recurrent neural network according to the extracted time sequence characteristics;
s206, decoding the output of the bidirectional recurrent neural network to obtain a final recognition result.
Due to the adoption of the technical scheme, the technical progress of the invention is as follows.
The cloud deck is arranged in the pipeline violation behavior identification system, the plugging conditions of two ends of the pipeline can be detected simultaneously through the cloud deck in the flight process of the unmanned aerial vehicle, and the problem that the pipeline plug of the unmanned aerial vehicle in the navigation direction can only be detected in the flight process of the unmanned aerial vehicle in the prior art is solved. The invention can meet the requirement of the unmanned aerial vehicle on line patrol of the pipeline plug in the horizontal direction and can also meet the requirement of the unmanned aerial vehicle on line patrol of the pipeline with an inclined angle. The invention can identify the data of the pipeline interface, identify whether workers have safety caps or not, detect the condition of people standing on the pipeline, directionally shoot the condition of the end plugs of the pipeline ports, judge the blocking condition of the pipeline and realize the effective supervision of the illegal behaviors of the pipeline.
Drawings
FIG. 1 is a schematic flow chart of the method for preprocessing, filtering and identifying images according to the present invention.
Detailed Description
The invention will be described in further detail below with reference to the figures and specific examples.
Pipeline act of violating regulations identification system includes unmanned aerial vehicle, camera, grey level camera, range finding sensor and cloud platform.
The unmanned aerial vehicle is a carrier of the identification system; be provided with GPS position sensor on the unmanned aerial vehicle, can control unmanned aerial vehicle through the cloud platform and fly to the assigned position.
The camera sets up on unmanned aerial vehicle, can realize the rotation of six degrees of freedom.
The gray camera is arranged on the unmanned aerial vehicle, is not provided with a holder, and can process data and method of the interface in real time on line.
And the distance measuring sensor is arranged at the lower end of the holder and can detect and judge the inclined pipeline in the flying process. Through the combination of the fixed gray level camera and the distance measuring sensor, key areas in pipeline detection can be extracted.
The cloud platform sets up on the camera, can detect the shutoff condition at pipeline both ends simultaneously through the cloud platform at unmanned aerial vehicle flight in-process, has solved among the prior art problem that only can detect the pipeline end cap of unmanned aerial vehicle navigation direction at unmanned aerial vehicle flight in-process. Under the angle control of the holder, the camera is controlled to rotate, and then the condition of the plug of the pipeline is detected. Simultaneously, can adjust unmanned aerial vehicle's height through the cloud platform, set up unmanned aerial vehicle's flying speed, carry out accurate image and shoot, play the effect of safe supervision.
The invention can meet the requirement of the unmanned aerial vehicle on line patrol of the pipeline plug in the horizontal direction and can also meet the requirement of the unmanned aerial vehicle on line patrol of the pipeline with an inclined angle.
The pipeline violation behavior identification method comprises the following steps:
and S1, pre-flying of the unmanned aerial vehicle. The method mainly comprises the steps of collecting pipeline data through a camera in the process of pre-flying of the unmanned aerial vehicle, preprocessing images, filtering out data which are not available on the periphery of the pipeline through a gray level camera by adopting an image filtering method, identifying pipeline interface data through a pipeline port identification method, identifying whether workers have safety helmets or not through a method that constructors do not carry the safety helmets, detecting the situations of people who stand on the pipeline through a pipeline people-standing detection method, and extracting key areas in pipeline detection.
The method for preprocessing, filtering and identifying the image, disclosed by the invention, is combined with the method shown in figure 1, and comprises the following steps of:
A. and (3) fuzzy image processing: preprocessing and block matching are carried out on the aerial photo, image blocks of RGB three channel components are searched, and image blocks similar to the image blocks in the city block are searched;
matrix recovery is carried out on a low-rank matrix composed of similar image blocks, and the purpose of image denoising is achieved;
estimating a fuzzy kernel by utilizing normalized sparse prior;
and recovering a clear image from the blurred image by using a non-blind deblurring method.
B. Image segmentation: carrying out graying processing on the processed picture;
and obtaining superpixels which are uniformly distributed and attached to the edge by using an SLIC superpixel segmentation algorithm, realizing image segmentation, further performing edge optimization on the image, and better recovering the edge information of the original image.
C. Image filtering: and processing the noise of the divided picture by utilizing the adaptive median wave construction.
D. Image recognition: and obtaining a differential image by using a background image modeling method, judging the difference between the target image and the original image, and further judging whether the pipeline stands or has foreign matters.
Detecting the condition of people on the pipeline by a method for judging people on the pipeline; the method for distinguishing the person standing on the pipeline comprises the following steps:
s101, firstly, collecting images of a plurality of frames of video without foreign matters within a period of time, summing the images and then averaging the images to be used as an initial background image, namely
Wherein, BI(x, y) is the initial background image, Fi(x, y) represents the ith frame image.
S102, obtaining a difference image by the difference between the current frame image and the background image, carrying out binarization processing on the difference image, and preliminarily separating the target from the background.
S103, filtering isolated foreground points in the binary image through morphological corrosion operation, and performing expansion operation to obtain a complete target outline, so as to determine whether foreign matters exist on the pipeline or whether dangerous behaviors of people standing on the pipeline exist.
Judging whether the personnel come by a personnel identification method, then carrying out binarization processing, and judging whether the personnel have safety helmets or not by the top color of the heads.
The method for identifying whether a worker carries a safety helmet comprises the following steps:
s111, a YOLOv4-tiny lightweight neural network is utilized, and a backbone network of the YOLOv4-tiny lightweight neural network mainly comprises a downsampling CBL structure and a CSP structure. The underlying features can provide more accurate location information, and the use of the max-pooling layer can reduce image size and extract key information.
And simultaneously, an MsxModule structure is constructed and added in a multi-scale fusion process to extract the main characteristics of small and medium-sized targets, so that the error of target positioning of a deep network caused by a series of sampling CBL structures and a plurality of convolution operations is eliminated.
The transmission path from the shallow layer feature to the deep layer feature is long, the edge information and the positioning information of the shallow layer feature are easy to lose, the data utilization rate is low, the detection precision is not ideal, and the like, and in order to fully use the feature information, a multi-scale feature fusion structure from bottom to top is used for a branch network outputting a large detection head.
S112, more main features can be obtained by utilizing a Max Module structure, the edge information and the positioning information of the shallow network are reserved by a bottom-up feature fusion structure, the feature utilization rate is improved, and a feedforward transfer equation is as follows:
the weight update equation is:
and (3) positions of the CIoU prediction frame and the real frame are quoted, so that the loss convergence speed is accelerated, and the loss function is as follows:
s113, building a data set of the wearable safety helmet, using k-means + + to reconstruct the proportion of the anchor frame, and processing the training set by adopting a Mosaic method, so that the generalization capability of the model in an actual detection scene is improved, and the model is more suitable for the detection scene of the wearable safety helmet.
S2, after the first pre-flying, the unmanned aerial vehicle is positioned through a GPS position sensor arranged on the unmanned aerial vehicle, flies to a pipeline key area through the control of a holder, identifies a pipeline port, controls a camera of the unmanned aerial vehicle to adjust the direction through control software, directionally shoots the condition of a pipeline port plug, and judges the plugging condition of the pipeline through a pipeline port plugging method.
In the process of identifying the pipeline port, the labels are set to be two types, one type is that the pipeline is blocked, and the other type is that the pipeline is not blocked, so that the labels with only the two types of results can be matched through the picture transmitted by the camera. The method is similar to the principle of license plate recognition or traffic sign recognition, and can be matched in a plurality of generated sequence labels in license plate recognition.
Step S2 includes the following steps:
s201, directly positioning the pipeline port by using Yolov 3.
And S202, effectively correcting the positioned inclined port by using the STN network.
S203, improving the CNN network; two residual error network structures are added into the CNN network, so that the network bracelet is faster and gradient disappearance is not easy to occur.
And S204, extracting time sequence characteristics by using the improved CNN network and the corrected picture.
And S205, identifying the time sequence label by adopting a Bidirectional Recurrent Neural Network (BRNN) according to the extracted time sequence characteristics.
S206, decoding the output of the Bidirectional Recurrent Neural Network (BRNN) by adopting a time sequence classification network CTC to obtain a final identification result.
The invention utilizes the Faster R-CNN network to train the data set, and adopts 70 percent as a training set, 15 percent as a testing set and 15 percent as a verification set. After the data set is determined, the images in the training set are labeled, so that the network learns the characteristics of whether the pipeline is blocked or not, and the classification is carried out better.
In order to obtain the highest accuracy of detection, a Classifier enhanced network model is used to obtain a loss function.
Claims (5)
1. The pipeline violation behavior identification method is characterized by comprising the following steps:
s1, pre-flying of the unmanned aerial vehicle; collecting pipeline data in the process of pre-flying of the unmanned aerial vehicle, preprocessing images, filtering data which is not available on the periphery of a pipeline, identifying pipeline interface data, identifying whether workers have safety helmets, detecting the condition of people standing on the pipeline, and extracting key areas in pipeline detection;
s2, controlling the unmanned aerial vehicle to fly to a key area, identifying a pipeline port, shooting the condition of a pipeline port plug in an oriented mode, and judging the plugging condition of the pipeline.
2. The pipeline violation behavior recognition method of claim 1, wherein the condition of the pipeline person is detected by a pipeline person-on-station discrimination method; the method for distinguishing the people standing on the pipeline comprises the following steps:
s101, collecting images of a plurality of frames of video without foreign matters within a period of time, summing the images and averaging the images to obtain an initial background image, namely the initial background image
Wherein, BI(x, y) is the initial background image, Fi(x, y) represents the ith frame image;
s102, obtaining a difference image by subtracting the current frame image from the background image, carrying out binarization processing on the difference image, and preliminarily separating a target from the background;
s103, filtering isolated foreground points in the binary image, performing expansion operation to obtain a complete target profile, and determining whether foreign matters exist on the pipeline or whether dangerous behaviors of people standing on the pipeline exist.
3. The pipeline violation identification method of claim 1, wherein the personnel is identified by a personnel identification method, and then binarization processing is performed, and the personnel is identified by overhead color whether a safety helmet is provided.
4. The pipeline violation identification method of claim 3 wherein the method of identifying whether a worker has a safety helmet comprises the steps of:
s111, constructing an MsxModule structure by using a YOLOv4-tiny lightweight neural network, and adding the MsxModule structure in a multi-scale fusion process to extract main characteristics of small and medium-sized targets;
s112, acquiring more main features by utilizing a Max Module structure, reserving edge information and positioning information of the shallow layer network by utilizing a bottom-up feature fusion structure, wherein a feedforward transfer equation is as follows:
the weight update equation is:
and (3) positions of the CIoU prediction frame and the real frame are quoted, so that the loss convergence speed is accelerated, and the loss function is as follows:
s113, constructing a data set of the wearable safety helmet, reconstructing the proportion of the anchor frames, and processing a training set.
5. The pipeline violation identification method of claim 1, wherein said step S2 comprises the steps of:
s201, directly positioning a pipeline port;
s202, effectively correcting the positioned inclined port;
s203, improving the CNN network;
s204, extracting time sequence characteristics by using the improved CNN network and the corrected picture;
s205, identifying the time sequence label by adopting a bidirectional recurrent neural network according to the extracted time sequence characteristics;
s206, decoding the output of the bidirectional recurrent neural network to obtain a final recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110983603.8A CN113780113A (en) | 2021-08-25 | 2021-08-25 | Pipeline violation behavior identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110983603.8A CN113780113A (en) | 2021-08-25 | 2021-08-25 | Pipeline violation behavior identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113780113A true CN113780113A (en) | 2021-12-10 |
Family
ID=78839095
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110983603.8A Pending CN113780113A (en) | 2021-08-25 | 2021-08-25 | Pipeline violation behavior identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113780113A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513061A (en) * | 2015-12-02 | 2016-04-20 | 上海海事大学 | Method for automatically searching and rescuing person in distress on sea through employing unmanned plane |
| CN109379564A (en) * | 2018-10-30 | 2019-02-22 | 长春市万易科技有限公司 | A kind of gas pipeline unmanned plane inspection device and method for inspecting |
| CN110135267A (en) * | 2019-04-17 | 2019-08-16 | 电子科技大学 | A kind of subtle object detection method of large scene SAR image |
| CN110349189A (en) * | 2019-05-31 | 2019-10-18 | 广州铁路职业技术学院（广州铁路机械学校） | A kind of background image update method based on continuous inter-frame difference |
| CN110414359A (en) * | 2019-07-01 | 2019-11-05 | 中国石化销售有限公司华南分公司 | The analysis of long distance pipeline unmanned plane inspection data and management method and system |
| CN111339893A (en) * | 2020-02-21 | 2020-06-26 | 哈尔滨工业大学 | Pipeline detection system and method based on deep learning and unmanned aerial vehicle |
| CN111582272A (en) * | 2020-04-30 | 2020-08-25 | 平安科技（深圳）有限公司 | Double-row license plate recognition method, device and equipment and computer readable storage medium |
| KR102167414B1 (en) * | 2020-04-27 | 2020-10-19 | 이한구 | System and method for providing traffic violation detecting service using drone |
-
2021
- 2021-08-25 CN CN202110983603.8A patent/CN113780113A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513061A (en) * | 2015-12-02 | 2016-04-20 | 上海海事大学 | Method for automatically searching and rescuing person in distress on sea through employing unmanned plane |
| CN109379564A (en) * | 2018-10-30 | 2019-02-22 | 长春市万易科技有限公司 | A kind of gas pipeline unmanned plane inspection device and method for inspecting |
| CN110135267A (en) * | 2019-04-17 | 2019-08-16 | 电子科技大学 | A kind of subtle object detection method of large scene SAR image |
| CN110349189A (en) * | 2019-05-31 | 2019-10-18 | 广州铁路职业技术学院（广州铁路机械学校） | A kind of background image update method based on continuous inter-frame difference |
| CN110414359A (en) * | 2019-07-01 | 2019-11-05 | 中国石化销售有限公司华南分公司 | The analysis of long distance pipeline unmanned plane inspection data and management method and system |
| CN111339893A (en) * | 2020-02-21 | 2020-06-26 | 哈尔滨工业大学 | Pipeline detection system and method based on deep learning and unmanned aerial vehicle |
| KR102167414B1 (en) * | 2020-04-27 | 2020-10-19 | 이한구 | System and method for providing traffic violation detecting service using drone |
| CN111582272A (en) * | 2020-04-30 | 2020-08-25 | 平安科技（深圳）有限公司 | Double-row license plate recognition method, device and equipment and computer readable storage medium |
Non-Patent Citations (5)
| Title |
|---|
| 吴志强主编: "《智能城市》", 30 September 2020, 上海科学技术出版社, pages: 147 - 150 * |
| 李道亮编著: "《无人农场》", 31 August 2020, 机械工业出版社, pages: 229 - 231 * |
| 王兵 等: "改进YOLO轻量化网络的口罩检测算法", 计算机工程与应用, vol. 57, no. 8, 15 April 2021 (2021-04-15), pages 62 - 69 * |
| 王兵: "改进YOLO轻量化网络的口罩检测算法", 计算机工程与应用, vol. 57, no. 08, 19 January 2021 (2021-01-19), pages 62 - 69 * |
| 甘胜丰，雷维新，周成宏著: "《机器视觉表面缺陷检测技术及其在钢铁工业中的应用》", 30 June 2017, 华中科技大学出版社, pages: 84 - 85 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105373135B (en) | A kind of method and system of aircraft docking guidance and plane type recognition based on machine vision | |
| CN103106766B (en) | Forest fire identification method and forest fire identification system | |
| CN101694681B (en) | Bird striking risk assessment system and assessment method thereof | |
| CN104933680B (en) | A kind of intelligent quick sea fog minimizing technology of unmanned boat vision system video | |
| CN106816038A (en) | A kind of inland waters abnormal behaviour ship automatic identification system and method | |
| CN104766086A (en) | Supervising method and system of way mark | |
| CN111241959A (en) | Method for detecting person without wearing safety helmet through construction site video stream | |
| CN103632158A (en) | Forest fire prevention monitor method and forest fire prevention monitor system | |
| CN113516076A (en) | Improved lightweight YOLO v4 safety protection detection method based on attention mechanism | |
| CN101145200A (en) | Inner river ship automatic identification system of multiple vision sensor information fusion | |
| CN102136076A (en) | Method for positioning and tracing underground personnel of coal mine based on safety helmet detection | |
| CN107729866A (en) | Ship based on timing diagram picture touches mark automatic detection device and method | |
| CN108509919A (en) | A kind of detection and recognition methods based on deep learning to waterline in video or picture | |
| CN114495068B (en) | Pavement health detection method based on human-computer interaction and deep learning | |
| CN112417931A (en) | Method for detecting and classifying water surface objects based on visual saliency | |
| CN113537019A (en) | Detection method for identifying wearing of safety helmet of transformer substation personnel based on key points | |
| CN108563986A (en) | Earthquake region electric pole posture judgment method based on wide-long shot image and system | |
| CN106372566A (en) | Digital signage-based emergency evacuation system and method | |
| CN109325911B (en) | Empty base rail detection method based on attention enhancement mechanism | |
| CN105335985B (en) | A kind of real-time capturing method and system of docking aircraft based on machine vision | |
| CN113780113A (en) | Pipeline violation behavior identification method | |
| CN113989886B (en) | Crewman identity verification method based on face recognition | |
| CN115690743A (en) | Airport gate stand intrusion detection method based on deep learning | |
| Bouhsine et al. | Atmospheric visibility image-based system for instrument meteorological conditions estimation: A deep learning approach | |
| CN107240111A (en) | Edge connection segmentation passenger flow statistical method |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Dai bingtao Inventor after: Li Yong Inventor after: Han Mingyi Inventor after: Zhao Junwei Inventor after: Xia Rongbei Inventor before: Dai bingtao |
|
| CB03 | Change of inventor or designer information |