CN117994220A - Highway pavement detection method - Google Patents

Highway pavement detection method Download PDF

Info

Publication number
CN117994220A
CN117994220A CN202410129893.3A CN202410129893A CN117994220A CN 117994220 A CN117994220 A CN 117994220A CN 202410129893 A CN202410129893 A CN 202410129893A CN 117994220 A CN117994220 A CN 117994220A
Authority
CN
China
Prior art keywords
road surface
image
highway
pavement
detecting
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
Application number
CN202410129893.3A
Other languages
Chinese (zh)
Inventor
李捷飞
黄伟
杨建湖
张杰文
曹剑清
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yunnan Yunlu Engineering Testing Co ltd
Original Assignee
Yunnan Yunlu Engineering Testing Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Yunnan Yunlu Engineering Testing Co ltd filed Critical Yunnan Yunlu Engineering Testing Co ltd
Priority to CN202410129893.3A priority Critical patent/CN117994220A/en
Publication of CN117994220A publication Critical patent/CN117994220A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B31/00—Predictive alarm systems characterised by extrapolation or other computation using updated historic data
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • 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
    • G06N3/088—Non-supervised learning, e.g. competitive learning
    • 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
    • G06N3/0895—Weakly supervised learning, e.g. semi-supervised or self-supervised learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/18—Status alarms
    • G08B21/182—Level alarms, e.g. alarms responsive to variables exceeding a threshold
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • Analytical Chemistry (AREA)
  • Pathology (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Chemical & Material Sciences (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to the technical field of road surface detection, and discloses a road surface detection method, which comprises the following steps: collecting pavement images and real-time data, carrying out image graying, denoising and binarization processing, carrying out data preprocessing, extracting characteristics of the images and the data, training a deep learning model and a decision model, inputting new images, calibrating and calibrating, triggering early warning below a threshold value, setting the threshold value according to the grading, and reminding to check or overhaul the pavement; compared with the traditional method, the method has higher accuracy and robustness, and can timely find the road surface problem, improve the efficiency and quality of road maintenance, further predict the future condition of the road, assist maintenance personnel to maintain the road surface, and timely early warn, timely find the road surface problem and provide the efficiency and quality of road maintenance by integrating technical means such as multi-source data, real-time monitoring and early warning systems and the like.

Description

Highway pavement detection method
Technical Field
The invention relates to the technical field of road surface detection, in particular to a road surface detection method.
Background
With the rapid development of highway traffic, the safety and durability of highway pavement are increasingly concerned, so that highway pavement detection becomes a key link for highway maintenance and management.
The traditional road surface detection method generally adopts a method based on threshold values or manual characteristics to evaluate the road surface condition, the methods depend on fixed parameters and experience, are difficult to adapt to different road conditions and environmental changes, and are easy to generate misjudgment, so that a monitoring method by means of visual images appears, and the method can detect the current road surface and know the road surface defect, but has certain defects, namely the future condition of the road surface cannot be predicted, the road surface problem cannot be found in time, and the efficiency and quality of road maintenance are affected.
Disclosure of Invention
The invention provides a highway pavement detection method for solving the problems in the background art.
In order to achieve the above object, the present invention provides the following technical solutions: a highway pavement detection method comprises the following steps:
s1: collecting road surface images through shooting equipment, and collecting road surface data in real time through a sensor;
S2: carrying out graying, denoising and binarization operation on the acquired image, optimizing the image quality, and preprocessing pavement data monitored in real time;
s3: extracting characteristics such as textures, cracks, pits and the like in the pavement image through a preset algorithm, and simultaneously extracting characteristics of the preprocessed real-time monitoring data, wherein the characteristics comprise abnormal temperature and abnormal humidity;
s4: training a deep learning model by extracting the characteristics of the road surface image, and simultaneously establishing a decision model by combining an enhanced learning algorithm;
S5: inputting the new road surface image into a trained deep learning model, obtaining a road surface condition score through automatic calibration and calibration functions, automatically triggering an early warning signal when the score is lower than a preset threshold value, and sending the early warning signal to maintenance personnel;
S6: according to the road surface condition scoring and decision model, setting a threshold A, B, wherein A is more than B, reminding a person to check the road surface when the scoring is lower than the threshold B, and reminding the person to overhaul the road surface when the scoring is lower than the threshold A.
Preferably, the deep learning model includes a convolutional neural network and a recurrent neural network.
Preferably, the preset algorithm comprises a Sobel operator and a Canny edge detection algorithm.
Preferably, the road surface data includes temperature, humidity and vehicle flow.
Preferably, the photographing apparatus includes a high-resolution camera or a camera carried by an unmanned aerial vehicle.
Preferably, the formula of graying is:
Y＝0.2989*R+0.5870*G+0.1140*B；
Wherein Y represents a gray scale value, R represents a red channel luminance value of an image, G represents a green channel luminance value, and B represents a blue channel luminance value.
Preferably, the formula of denoising is:
(I_{filtered}＝I_{original}*h)；
Where (I_ { filtered }) is the denoised image, (I_ { original }) is the original image, and h is a Gaussian filter.
Preferably, the computational formula of the Canny edge detection is as follows:
[E＝\sqrt{d^2_1+d^2_2}][T＝\frac{\sigma}{2}*max(E)]；
Wherein E is the gradient amplitude, d1 and d2 are the gradient directions of two pixel points adjacent to the edge point, sigma is the standard deviation of the Gaussian filter, and T is the threshold.
Preferably, the formula of the convolutional neural network is:
[y＝f(W*x+b)]；
Where y is the output, f is the activation function (e.g., reLU), W is the convolution kernel weight matrix, b is the bias vector, and x is the input feature matrix.
Preferably, the formula of the recurrent neural network is:
[h_t＝f(W_h*x_t+U*h_{t-1}+b_h)]；
wherein h_t is a hidden state at the current moment, x_t is an input vector at the current moment, w_h and U are respectively a weight matrix and a transfer matrix, and b_h is a bias vector.
In the technical scheme, the invention has the technical effects and advantages that:
The invention adopts the deep learning technology to predict the road surface condition, has higher accuracy and robustness compared with the traditional method, can timely discover the road surface problem by integrating technical means such as multi-source data, a real-time monitoring and early warning system and the like, improves the efficiency and quality of road maintenance, further can predict the future condition of the road, assists maintenance personnel to maintain the road surface, can realize real-time early warning, can automatically trigger corresponding maintenance measures when the score is lower than a threshold value, can timely discover the road surface problem, and provides the efficiency and quality of road maintenance.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without any inventive effort, are intended to be within the scope of the invention.
The technical solution in the embodiment of the present application is to solve the above problems, as shown in fig. 1, the general idea is as follows:
Example 1
In this embodiment, use the high-resolution camera as shooting equipment to select unmanned aerial vehicle to carry on the camera and carry out road surface image acquisition, through unmanned aerial vehicle's high altitude visual angle, can acquire more comprehensive road surface image, provide richer data for follow-up feature extraction and model training.
In the image preprocessing stage, the image is subjected to graying processing by adopting a graying formula y=0.2989r+0.5870g+0.1140×b to reduce the calculated amount and improve the processing speed, and meanwhile, the image is subjected to denoising processing by using a gaussian filter, wherein the formula is i_ { filtered } =i_ { original }.
And in the feature extraction stage, texture and crack features in the pavement image are extracted by utilizing a Sobel operator and a Canny edge detection algorithm, the Sobel operator detects the pavement edge by calculating the gradient strength of the pixel points in the horizontal and vertical directions, and the Canny edge detection algorithm detects cracks and other edge information by calculating the strength and the direction of the pixel points in the gradient directions.
In the deep learning model training stage, a Convolutional Neural Network (CNN) and a cyclic neural network (RNN) are selected for training, wherein the convolutional neural network is used for extracting local features in road surface images, and the cyclic neural network is used for processing sequence data such as road surface temperature, humidity, vehicle flow and the like.
In the road surface condition prediction stage, a new road surface image is input into a trained deep learning model, road surface condition scores are obtained through automatic calibration and calibration functions, and when the scores are lower than a preset threshold value, early warning signals are automatically triggered and sent to maintenance personnel.
In the pavement maintenance and repair stage, corresponding maintenance and repair schemes are formulated according to the grading of pavement conditions and the recommendation of a decision model, and when the grading is lower than a threshold value A, corresponding repair operation is triggered; when the score is below threshold B, the person is alerted to examine the road surface.
Example two
In this embodiment, more advanced deep learning models are used for road surface condition prediction, and in addition to convolutional neural networks and recurrent neural networks, advanced models such as attention mechanism networks (AttentionNetwork) and generation countermeasure networks (GAN) are introduced.
The attention mechanism network enables the model to pay attention to important areas in the pavement image better by introducing attention mechanisms into the model, so that the characteristics are extracted more accurately, and the countermeasure network is generated to generate more realistic pavement images for expanding a data set and improving the performance of the model.
In the feature extraction stage, more algorithms and technologies such as wavelet transformation and Fourier transformation are introduced to more comprehensively extract the features in the road surface image, and meanwhile, a feature fusion technology is used to fuse different features so as to improve the prediction accuracy of the model.
In order to further improve the accuracy and the instantaneity of early warning, the internet of things technology is introduced in the transmission of early warning signals, and the early warning signals are transmitted to nearby maintenance personnel or vehicles to realize quick response and processing.
In the pavement maintenance and repair stage, an intelligent inspection robot is adopted for pavement inspection and repair work, and the robot is provided with advanced sensors and actuators, so that pavement condition detection, evaluation and repair work can be automatically completed.
Example III
In the embodiment, the training process of the deep learning model is further optimized, an unsupervised learning algorithm and a semi-supervised learning algorithm are added besides the traditional supervised learning algorithm, and the unsupervised learning algorithm is used for learning the unlabeled data, so that potential characteristics and modes are extracted; and then, the label data is combined to perform semi-supervised learning, so that the generalization capability of the model is further improved.
In order to improve the robustness and stability of the model, transfer learning and fine tuning techniques are introduced in the training process. The pre-training model is migrated to a specific task and is subjected to fine adjustment, so that the model is better adapted to the road surface condition prediction task under a specific scene.
Example IV
A highway pavement detection method comprises the following steps:
s1: the method comprises the steps that road surface images are collected through shooting equipment, the shooting equipment comprises a high-resolution camera or a camera carried by an unmanned aerial vehicle, and road surface data are collected through a sensor in real time;
S2: carrying out graying, denoising and binarization operation on the acquired image, optimizing the image quality, and preprocessing pavement data monitored in real time, wherein the pavement data comprises temperature, humidity and traffic flow;
s3: extracting characteristics such as textures, cracks, pits and the like in the pavement image through a preset algorithm, and simultaneously extracting characteristics of the preprocessed real-time monitoring data, wherein the characteristics comprise abnormal temperature and abnormal humidity;
S4: training a deep learning model by extracting features of a pavement image, and simultaneously establishing a decision model by combining an enhanced learning algorithm
S5: inputting the new road surface image into a trained deep learning model, obtaining a road surface condition score through automatic calibration and calibration functions, automatically triggering an early warning signal when the score is lower than a preset threshold value, and sending the early warning signal to maintenance personnel;
S6: according to the road surface condition scoring and decision model, setting a threshold A, B, wherein A is more than B, reminding a person to check the road surface when the scoring is lower than the threshold B, and reminding the person to overhaul the road surface when the scoring is lower than the threshold A.
In addition, in the present invention, the deep learning model includes a convolutional neural network and a cyclic neural network.
And, regarding the above preset algorithm, the preset algorithm includes Sobel operator and Canny edge detection algorithm.
For further explanation of the above, regarding the above graying, the formula of graying is:
Y＝0.2989*R+0.5870*G+0.1140*B；
Wherein Y represents a gray scale value, R represents a red channel luminance value of an image, G represents a green channel luminance value, and B represents a blue channel luminance value.
For further explanation of the above, regarding the above denoising, the formula of denoising is:
(I_{filtered}＝I_{original}*h)；
Where (I_ { filtered }) is the denoised image, (I_ { original }) is the original image, and h is a Gaussian filter.
For further explanation of the above, regarding the above Canny edge detection, the Canny edge detection formula is:
The computational formula of Canny edge detection is:
[E＝\sqrt{d^2_1+d^2_2}][T＝\frac{\sigma}{2}*max(E)]；
Wherein E is the gradient amplitude, d1 and d2 are the gradient directions of two pixel points adjacent to the edge point, sigma is the standard deviation of the Gaussian filter, and T is the threshold.
For further explanation of the above, regarding the convolutional neural network, the formula of the convolutional neural network is:
[y＝f(W*x+b)]；
Where y is the output, f is the activation function (e.g., reLU), W is the convolution kernel weight matrix, b is the bias vector, and x is the input feature matrix.
For further explanation of the above, the formula for the recurrent neural network is:
the formula of the recurrent neural network is:
[h_t＝f(W_h*x_t+U*h_{t-1}+b_h)]；
wherein h_t is a hidden state at the current moment, x_t is an input vector at the current moment, w_h and U are respectively a weight matrix and a transfer matrix, and b_h is a bias vector.
While certain exemplary embodiments of the present invention have been described above by way of illustration only, it will be apparent to those of ordinary skill in the art that modifications may be made to the described embodiments in various different ways without departing from the spirit and scope of the invention. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive of the scope of the invention, which is defined by the appended claims.

Claims (10)

1. A highway pavement detection method is characterized in that: the method comprises the following steps:
s1: collecting road surface images through shooting equipment, and collecting road surface data in real time through a sensor;
S2: carrying out graying, denoising and binarization operation on the acquired image, optimizing the image quality, and preprocessing pavement data monitored in real time;
s3: extracting characteristics such as textures, cracks, pits and the like in the pavement image through a preset algorithm, and simultaneously extracting characteristics of the preprocessed real-time monitoring data, wherein the characteristics comprise abnormal temperature and abnormal humidity;
s4: training a deep learning model by extracting the characteristics of the road surface image, and simultaneously establishing a decision model by combining an enhanced learning algorithm;
S5: inputting the new road surface image into a trained deep learning model, obtaining a road surface condition score through automatic calibration and calibration functions, automatically triggering an early warning signal when the score is lower than a preset threshold value, and sending the early warning signal to maintenance personnel;
S6: according to the road surface condition scoring and decision model, setting a threshold A, B, wherein A is more than B, reminding a person to check the road surface when the scoring is lower than the threshold B, and reminding the person to overhaul the road surface when the scoring is lower than the threshold A.
2. The method for detecting a highway surface according to claim 1, wherein: the deep learning model includes a convolutional neural network and a recurrent neural network.
3. The method for detecting a highway pavement according to claim 2, wherein: the preset algorithm comprises a Sobel operator and a Canny edge detection algorithm.
4. A method of highway pavement detection according to claim 3 and wherein: the road surface data comprises temperature, humidity and vehicle flow.
5. The method for detecting a highway surface according to claim 4, wherein: the shooting equipment comprises a high-resolution camera or a camera carried by the unmanned aerial vehicle.
6. The method for detecting a highway surface according to claim 5, wherein: the graying formula is as follows:
Y＝0.2989*R+0.5870*G+0.1140*B；
Wherein Y represents a gray scale value, R represents a red channel luminance value of an image, G represents a green channel luminance value, and B represents a blue channel luminance value.
7. The method for detecting a highway surface according to claim 6, wherein: the denoising formula is as follows:
(I_{filtered}＝I_{original}*h)；
Where (I_ { filtered }) is the denoised image, (I_ { original }) is the original image, and h is a Gaussian filter.
8. The method for detecting a highway surface according to claim 7, wherein: the computational formula of the Canny edge detection is as follows:
[E＝\sqrt{d^2_1+d^2_2}][T＝\frac{\sigma}{2}*max(E)]；
Wherein E is the gradient amplitude, d1 and d2 are the gradient directions of two pixel points adjacent to the edge point, sigma is the standard deviation of the Gaussian filter, and T is the threshold.
9. The method for detecting a highway surface according to claim 8, wherein: the formula of the convolutional neural network is as follows:
[y＝f(W*x+b)]；
Where y is the output, f is the activation function (e.g., reLU), W is the convolution kernel weight matrix, b is the bias vector, and x is the input feature matrix.
10. The method for detecting a highway surface according to claim 9, wherein: the formula of the cyclic neural network is as follows:
[h_t＝f(W_h*x_t+U*h_{t-1}+b_h)]；
wherein h_t is a hidden state at the current moment, x_t is an input vector at the current moment, w_h and U are respectively a weight matrix and a transfer matrix, and b_h is a bias vector.
CN202410129893.3A 2024-01-30 2024-01-30 Highway pavement detection method Pending CN117994220A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410129893.3A CN117994220A (en) 2024-01-30 2024-01-30 Highway pavement detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410129893.3A CN117994220A (en) 2024-01-30 2024-01-30 Highway pavement detection method

Publications (1)

Publication Number Publication Date
CN117994220A true CN117994220A (en) 2024-05-07

Family

ID=90900886

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410129893.3A Pending CN117994220A (en) 2024-01-30 2024-01-30 Highway pavement detection method

Country Status (1)

Country Link
CN (1) CN117994220A (en)

Similar Documents

Publication Publication Date Title
CN110261436B (en) Rail fault detection method and system based on infrared thermal imaging and computer vision
CN110084165B (en) Intelligent identification and early warning method for abnormal events in open scene of power field based on edge calculation
WO2020199538A1 (en) Bridge key component disease early-warning system and method based on image monitoring data
CN101430195B (en) Method for computing electric power line ice-covering thickness by using video image processing technology
CN110232380A (en) Fire night scenes restored method based on Mask R-CNN neural network
CN105809679A (en) Mountain railway side slope rockfall detection method based on visual analysis
Pandey et al. Autopilot control unmanned aerial vehicle system for sewage defect detection using deep learning
Munawar Image and video processing for defect detection in key infrastructure
CN111881970A (en) Intelligent outer broken image identification method based on deep learning
CN111144301A (en) Road pavement defect quick early warning device based on degree of depth learning
CN115995058A (en) Power transmission channel safety on-line monitoring method based on artificial intelligence
CN115600124A (en) Subway tunnel inspection system and inspection method
CN115527170A (en) Method and system for identifying closing fault of door stopper handle of automatic freight car derailing brake device
Chu et al. Deep learning method to detect the road cracks and potholes for smart cities
CN113077423B (en) Laser selective melting pool image analysis system based on convolutional neural network
CN112329858B (en) Image recognition method for breakage fault of anti-loosening iron wire of railway motor car
KR102281100B1 (en) System and method for providing heat transporting pipe status information
Narlan et al. Automated pavement defect detection using YOLOv8 object detection algorithm
CN117197019A (en) Vehicle three-dimensional point cloud image fusion method and system
CN112967335A (en) Bubble size monitoring method and device
CN117994220A (en) Highway pavement detection method
CN106128105A (en) A kind of traffic intersection pedestrian behavior monitoring system
CN104408942A (en) Intelligent vehicle speed measuring device and method
Ramachandraiah et al. Evaluation of Pavement Surface Distress Using Image Processing and Artificial Neural Network
CN115861825B (en) 2C detection method based on image recognition

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication