CN115994901A - Automatic road disease detection method and system - Google Patents
Automatic road disease detection method and system Download PDFInfo
- Publication number
- CN115994901A CN115994901A CN202310101278.7A CN202310101278A CN115994901A CN 115994901 A CN115994901 A CN 115994901A CN 202310101278 A CN202310101278 A CN 202310101278A CN 115994901 A CN115994901 A CN 115994901A
- Authority
- CN
- China
- Prior art keywords
- road
- disease
- data
- disease detection
- diseases
- 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
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image recognition processing, in particular to an automatic road disease detection method and system. The method comprises the steps of collecting road data through a camera and a laser radar; screening the obtained road data and judging whether suspected road diseases exist or not; when suspected road diseases are screened, storing the road data with the diseases into a buffer area; obtaining a road disease detection and identification result in the deep learning identification model; correcting the disease detection and identification result; the image for identifying the road disease is added with GPS coordinates, road name and type of the road disease. The invention can realize automatic disease detection, and the detection personnel can obtain the road surface information only in the maintenance vehicle without manual intervention in the whole process, thereby greatly reducing the working intensity of the personnel.
Description
Technical Field
The invention relates to the technical field of image recognition processing, in particular to an automatic road disease detection method and system.
Background
As the investment costs of highway facilities increase year by year, highway maintenance work becomes a concern. After the road is built, the road is affected by factors such as climate, geological conditions, traffic, load and the like, and the road can be damaged to different degrees along with the increase of years, so that road maintenance departments need to detect and maintain the road regularly. At present, the detection of road diseases is mainly based on a manual detection method, workers work outdoors for a long time, the detection efficiency is low, the working environment is bad, and how to realize the automatic detection of road diseases is the main research content of road maintenance at present.
In the prior art, the road disease detection is mainly based on manual visual inspection, has low speed, is dangerous, affects traffic and is inaccurate, and particularly on expressways, the danger coefficient is higher, so that the mode is gradually eliminated. Along with the continuous progress of science and technology, the vehicle-mounted acquisition equipment with high speed and high precision is widely applied at present, the equipment is utilized to patrol roads, the apparent data of the roads are acquired, the image data are screened piece by professionals, meanwhile, disease areas are marked, and then the disease length or area is calculated uniformly, so that the breakage rate is calculated. The road inspection vehicle has the advantages that the problems of low speed, danger and the like during manual detection are avoided, but the huge image database is screened and marked manually, so that the defects of time and labor waste and low accuracy still exist. Thus, vision-based disease detection techniques have evolved. However, the complexity and diversity of the road surface image and the weak signal nature of the disease information make it difficult to identify the disease algorithm based on the vision, and the quantitative calculation of the disease area with weak signal is required, which is always a difficulty, so an automatic detection method and system for the road disease is needed.
Disclosure of Invention
In order to solve the above-mentioned problems, the present invention provides an automatic detection method and system for road diseases.
In a first aspect, the present invention provides a method for automatically detecting road diseases, which adopts the following technical scheme:
an automatic road disease detection method comprises the following steps:
collecting road data through a camera and a laser radar;
screening the obtained road data and judging whether suspected road diseases exist or not;
when suspected road diseases are screened, storing the road data with the diseases into a buffer area;
obtaining a road disease detection and identification result in the deep learning identification model;
correcting the disease detection and identification result;
the image for identifying the road disease is added with GPS coordinates, road name and type of the road disease.
Further, the road data are collected through the camera and the laser radar, including starting a road disease detection system, calibrating coordinates and directions by using the sensor, and collecting the road data through the camera and the laser radar.
Further, the road data is collected through the camera and the laser radar, and the method further comprises the step of synchronously obtaining the time stamp and the detection position.
Further, the step of screening the obtained road data to judge whether the suspected road disease exists or not includes the step of calculating whether the probability of the road disease is within a set threshold range to obtain whether the suspected road disease exists or not.
Further, in the recognition model through deep learning, the road disease detection recognition result is obtained, and the image segmentation is carried out on the disease road data through a deep learning algorithm.
Further, in the recognition model through deep learning, a road disease detection recognition result is obtained, and the method further comprises the step of confirming whether the segmented image is a disease road or not through a deep learning algorithm.
Further, the correction of the disease detection and identification result comprises target division and image expansion of the disease road image according to the identification result.
In a second aspect, an automatic road disease detection system includes:
the data acquisition module is configured to acquire road data through the camera and the laser radar;
the screening module is configured to screen the acquired road data and judge whether suspected road diseases exist or not;
the storage module is configured to store the road data of the suspected road diseases into the buffer area when the suspected road diseases are screened;
the recognition module is configured to obtain a road disease detection recognition result through deep learning in the recognition model;
a correction module configured to correct the disease detection recognition result;
and an attaching module configured to attach the GPS coordinates, the road name, and the type of the road disease to the image in which the road disease is recognized.
In a third aspect, the present invention provides a computer-readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the method for automatically detecting a road fault.
In a fourth aspect, the present invention provides a terminal device, including a processor and a computer readable storage medium, where the processor is configured to implement instructions; the computer readable storage medium is for storing a plurality of instructions adapted to be loaded by a processor and to perform the method of automatically detecting a road condition.
In summary, the invention has the following beneficial technical effects:
according to the technical scheme, the road disease can be effectively identified, the road data is acquired through the assistance of the camera and the laser radar, the road data is identified by using the depth model, the correction function is achieved, and the accuracy of road disease identification can be greatly improved.
Drawings
Fig. 1 is a schematic flow chart of an automatic road disease detection method according to embodiment 1 of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
Example 1
Referring to fig. 1, a road disease automatic detection method of the present embodiment includes:
collecting road data through a camera and a laser radar;
screening the obtained road data and judging whether suspected road diseases exist or not; when suspected road diseases are screened, storing the road data with the diseases into a buffer area; obtaining a road disease detection and identification result in the deep learning identification model; correcting the disease detection and identification result; the image for identifying the road disease is added with GPS coordinates, road name and type of the road disease. The road data are collected through the camera and the laser radar, the road disease detection system is started, the sensor is used for calibrating coordinates and directions, and the road data are collected through the camera and the laser radar. The road data are collected through the camera and the laser radar, and the method further comprises the step of synchronously acquiring the time stamp and the detection position. And screening the obtained road data to judge whether the suspected road disease exists or not, wherein the step of calculating whether the probability of the road disease is within a set threshold range or not is carried out to obtain whether the suspected road disease exists or not. And obtaining a road disease detection and identification result in the deep learning identification model, wherein the road disease detection and identification result comprises image segmentation of the disease road data through a deep learning algorithm. In the recognition model through the deep learning, a road disease detection recognition result is obtained, and whether the segmented image is a disease road or not is confirmed through a deep learning algorithm. And correcting the disease detection and identification result, wherein the correction comprises target splitting and image expansion of the disease road image according to the identification result.
The method specifically comprises the following steps:
s1, screening the obtained road data and judging whether suspected road diseases exist or not;
the image acquisition can be realized by installing equipment such as a camera, an on-board processor, a flat panel controller, a 4G wireless router and the like on the maintenance vehicle. The maintenance vehicle is provided with a high-definition vehicle-mounted camera, a vehicle-mounted processor, a flat panel controller and a high-speed 4G wireless router. The high-definition vehicle-mounted camera is used for acquiring real-time video and video; the vehicle-mounted processor is used for running disease detection software and obtaining a disease detection result; the panel controller is used for camera control, manual snapshot, detection result viewing and the like; the high-speed 4G wireless router is used for TCP communication and transmits video, image and disease data to the background. The camera and the laser radar acquire road data and synchronize the data through a time stamp and positioning information, the detected diseased road in the embodiment mainly comprises pits, swelling bags and ruts on the road, the camera adopted in the embodiment is an area array camera, the laser radar adopted in the embodiment is a 64-line laser radar, the wide-angle camera is additionally arranged right above the roof of a test vehicle, the laser radar is simultaneously installed and fixed through a bracket, the camera is installed on the central line of the junction of the roof and the windshield, the visual field is inclined by 3 degrees downwards, the camera and the laser radar are calibrated by using a checkerboard and a calibration plate, the positions of the camera are all origins of a local coordinate system, and the right front of the vehicle is in the X-axis direction. The obtained road data is screened, the similar disease image and the similar road sign image are screened from the initial road data, and then the similar disease image and the similar road sign image are further screened to obtain the target disease image and the target road sign image, so that repeated calculation when the adjacent frame images in the detection video detect the road disease or the road sign at the same position can be reduced.
S2, when suspected road diseases are screened, storing the road data of the diseases into a cache area; the laser radar performs preliminary screening on the road data to judge whether a suspected disease road exists or not; if the calculated road surface disease probability is within a certain judging threshold value, the road disease is considered to be absent and the current road data is discarded. When the laser radar preliminary screening finds that a suspected disease road exists, the system stores the suspected disease road data into the data cache area.
S3, obtaining a road disease detection and identification result in the deep learning identification model;
the deep learning algorithm module finds out the image data of the suspected disease road in the data buffer area and cuts out the image of the corresponding area of the suspected disease road; the deep learning algorithm module can output the related disease category, the region and the confidence coefficient in the image, the training sample is derived from 30000 pieces of various disease pictures acquired in advance, the picture sizes are uniform to 1920 x 1080, and the model can self-adjust internal parameters in the training process so as to adapt to the training set.
The deep learning recognition model adopted in the embodiment is a YOLO-v3 algorithm model. The YOLO-v3 is a target detection open source algorithm based on a dark learning computing framework, the dark is exquisite and strong, the source code of the dark is written by a C language and a CUDA (compute unified device architecture) bottom layer, the code structure is strict, the speed is high, the parallel operation function of a multi-core processor and a GPU (graphics processing unit) is fully exerted, and the characteristic of an algorithm model of the YOLO-v3 is perfectly reflected; meanwhile, the method has very high accuracy on objects with medium and small sizes, so that YOLO-v3 is selected for real-time detection of road diseases.
Yolo-v3 employs end-to-end detection on the predicted picture, dividing the entire picture into S x S regions, which would be detected by the corresponding network if the center of an object falls on a region. Each network has a prediction area, four coordinate parameters for each prediction, coordinates Tx, ty in the upper left corner, width and height Tw, th, and confidence, which are the products of logistic regression. Confidence is used to determine if the predictions tend to be ignored, and if not, a logistic regression of the multi-label classification is performed to label. The deep learning algorithm module sets two thresholds A1 and A2 (A1 < A2); in this embodiment, based on the working vehicle speed of 50kph in this embodiment, two thresholds A1 and A2 (A1 < A2) are set, and for the target with the confidence lower than A1, it is directly considered that the target is not a disease road, and for the target with the confidence higher than A2, it is directly considered that the target is a disease road, that is, for the target with the confidence between A1 and A2 appearing in a plurality of pictures that are consecutive in time, the threshold A2 gradually decreases as the number of accumulation times N of consecutive time increases.
S4, correcting the disease detection and identification result; when the disease detection and identification result is a road crack, correcting the disease detection and identification result to distinguish an unrepaired crack from a repaired crack, wherein the correction process comprises the following steps: performing target splitting on the image of the detected road crack according to a threshold T according to the recognition result of the deep learning recognition model; expanding the image after target splitting by adopting a 3×3 template; filtering the expanded image; marking the connected domain of the filtered image, and counting the width of the connected domain; and if the width of the connected domain is larger than the filtering threshold value, modifying the disease detection and identification result of the image into repaired cracks. Performing target splitting on the image of the detected road crack according to the threshold value according to the recognition result of the deep learning recognition model; expanding the image after target splitting by adopting a 3×3 template; filtering the expanded image; marking the connected domain of the filtered image, and counting the width of the connected domain; and if the width of the connected domain is larger than the filtering threshold value, modifying the disease detection and identification result of the image into repaired cracks. The deep learning module obtains the width w and the height h of the region where the crack is located, calculates the segmentation threshold value T in the region by a histogram statistical method, namely, the distribution of gray values in the region where the crack is located is counted by the histogram statistical method, and takes the gray value with the largest occurrence number as the threshold value of target segmentation.
S5, adding GPS coordinates, road names and types of road diseases to the image with the identified road diseases, and uploading the identification results so as to inform relevant maintenance units of maintenance of the road diseases.
In the process of implementing the method of this embodiment, the maintenance personnel should operate as follows:
1) After a maintenance person starts the vehicle, the camera and the industrial personal computer are automatically electrified, and automatically run detection software, and the camera obtains a maintenance pavement video in real time according to the running condition of the vehicle;
2) Obtaining a road disease identification result by the real-time video through the trained deep learning training weight;
3) Correcting the crack detection result by using a traditional video image processing technology to obtain a final recognition result;
4) And (3) superposing information such as GPS position, road name and the like when the diseases occur, uploading disease pictures, videos and messages to the background, and checking later by the nursing staff.
Example 2
The embodiment provides an automatic road disease detection system, comprising:
the data acquisition module is configured to acquire road data through the camera and the laser radar;
the screening module is configured to screen the acquired road data and judge whether suspected road diseases exist or not;
the storage module is configured to store the road data of the suspected road diseases into the buffer area when the suspected road diseases are screened;
the recognition module is configured to obtain a road disease detection recognition result through deep learning in the recognition model;
a correction module configured to correct the disease detection recognition result;
and an attaching module configured to attach the GPS coordinates, the road name, and the type of the road disease to the image in which the road disease is recognized.
In a third aspect, the present invention provides a computer-readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the method for automatically detecting a road fault.
In a fourth aspect, the present invention provides a terminal device, including a processor and a computer readable storage medium, where the processor is configured to implement instructions; the computer readable storage medium is for storing a plurality of instructions adapted to be loaded by a processor and to perform the method of automatically detecting a road condition.
The above embodiments are not intended to limit the scope of the present invention, so: all equivalent changes in structure, shape and principle of the invention should be covered in the scope of protection of the invention.
Claims (8)
1. An automatic road disease detection method is characterized by comprising the following steps:
collecting road data through a camera and a laser radar;
screening the obtained road data and judging whether suspected road diseases exist or not;
when suspected road diseases are screened, storing the road data with the diseases into a buffer area;
obtaining a road disease detection and identification result in the deep learning identification model;
correcting the disease detection and identification result;
the image for identifying the road disease is added with GPS coordinates, road name and type of the road disease.
2. The automatic road disease detection method according to claim 1, wherein the step of collecting road data through the camera and the laser radar comprises starting a road disease detection system, calibrating coordinates and directions by using the sensor, and collecting the road data through the camera and the laser radar.
3. The automatic road disease detection method according to claim 2, wherein the collecting of road data by the camera and the lidar further comprises synchronously acquiring a time stamp and a detection position.
4. The method for automatically detecting road diseases according to claim 3, wherein the step of screening the obtained road data to determine whether a suspected road disease exists comprises calculating whether the probability of the road disease is within a set threshold range to obtain whether the suspected road disease exists.
5. The method according to claim 4, wherein obtaining the road disease detection and identification result from the deep learning identification model includes image segmentation of the disease road data by a deep learning algorithm.
6. The method according to claim 5, wherein the obtaining of the road disease detection recognition result by the deep learning recognition model further comprises confirming whether the segmented image is a damaged road by a deep learning algorithm.
7. The method according to claim 6, wherein correcting the disease detection recognition result includes performing object division and image expansion on the disease road image based on the recognition result.
8. An automatic road disease detection system, comprising:
the data acquisition module is configured to acquire road data through the camera and the laser radar;
the screening module is configured to screen the acquired road data and judge whether suspected road diseases exist or not;
the storage module is configured to store the road data of the suspected road diseases into the buffer area when the suspected road diseases are screened;
the recognition module is configured to obtain a road disease detection recognition result through deep learning in the recognition model;
a correction module configured to correct the disease detection recognition result;
and an attaching module configured to attach the GPS coordinates, the road name, and the type of the road disease to the image in which the road disease is recognized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310101278.7A CN115994901A (en) | 2023-02-13 | 2023-02-13 | Automatic road disease detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310101278.7A CN115994901A (en) | 2023-02-13 | 2023-02-13 | Automatic road disease detection method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115994901A true CN115994901A (en) | 2023-04-21 |
Family
ID=85990179
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310101278.7A Pending CN115994901A (en) | 2023-02-13 | 2023-02-13 | Automatic road disease detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115994901A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095316A (en) * | 2023-10-18 | 2023-11-21 | 深圳市思友科技有限公司 | Road surface inspection method, device, equipment and readable storage medium |
| CN117994225A (en) * | 2024-01-30 | 2024-05-07 | 中交第二公路勘察设计研究院有限公司 | Method and system for processing periodic detection data of multi-element pavement |
-
2023
- 2023-02-13 CN CN202310101278.7A patent/CN115994901A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095316A (en) * | 2023-10-18 | 2023-11-21 | 深圳市思友科技有限公司 | Road surface inspection method, device, equipment and readable storage medium |
| CN117095316B (en) * | 2023-10-18 | 2024-02-09 | 深圳市思友科技有限公司 | Road surface inspection method, device, equipment and readable storage medium |
| CN117994225A (en) * | 2024-01-30 | 2024-05-07 | 中交第二公路勘察设计研究院有限公司 | Method and system for processing periodic detection data of multi-element pavement |
| CN117994225B (en) * | 2024-01-30 | 2024-08-13 | 中交第二公路勘察设计研究院有限公司 | Method and system for processing periodic detection data of multi-element pavement |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Varadharajan et al. | Vision for road inspection | |
| CN111784657A (en) | Digital image-based system and method for automatically identifying cement pavement diseases | |
| CN112800911B (en) | Pavement damage rapid detection and natural data set construction method | |
| CN102073846B (en) | Method for acquiring traffic information based on aerial images | |
| CN115994901A (en) | Automatic road disease detection method and system | |
| CN114998852A (en) | Intelligent detection method for road pavement diseases based on deep learning | |
| CN114359181B (en) | Intelligent traffic target fusion detection method and system based on image and point cloud | |
| CN109782364B (en) | Traffic sign board missing detection method based on machine vision | |
| CN107798293A (en) | A kind of crack on road detection means | |
| CN113781537B (en) | Rail elastic strip fastener defect identification method and device and computer equipment | |
| CN111144301A (en) | Road pavement defect quick early warning device based on degree of depth learning | |
| CN113962960A (en) | Pavement disease detection method based on deep learning | |
| CN116863274A (en) | Semi-supervised learning-based steel plate surface defect detection method and system | |
| CN110610153A (en) | Lane recognition method and system for automatic driving | |
| CN115995056A (en) | Automatic bridge disease identification method based on deep learning | |
| Mertz et al. | City-wide road distress monitoring with smartphones | |
| CN116109616A (en) | Pavement crack detection and small-surface element fitting detection method based on YOLOv5 | |
| CN114972177A (en) | Road disease identification management method and device and intelligent terminal | |
| CN113312987B (en) | Recognition method based on unmanned aerial vehicle road surface crack image | |
| Yin et al. | Promoting Automatic Detection of Road Damage: A High-Resolution Dataset, a New Approach, and a New Evaluation Criterion | |
| CN112817006B (en) | Vehicle-mounted intelligent road disease detection method and system | |
| CN117095316B (en) | Road surface inspection method, device, equipment and readable storage medium | |
| CN113283309A (en) | Method and system for identifying road construction state according to real-time pavement image | |
| CN113762247A (en) | Road crack automatic detection method based on significant instance segmentation algorithm | |
| CN116229396B (en) | High-speed pavement disease identification and warning 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 |