Intelligent road inspection method and system and storage medium

Abstract

The invention discloses an intelligent road inspection method and system and a storage medium. The intelligent road inspection method comprises the steps that inspection task information issued by a first user and inspection service data of a second user are received; based on a pre-trained intelligent matching model, performing matching calculation on the inspection task information and the inspection service data to obtain a second user matched with the inspection task; issuing the inspection task to a matched second user, so that the matched second user automatically identifies road disease information based on a road intelligent identification module; and receiving inspection progress videos, images, sensor data and road disease identification information which are uploaded in real time in the inspection process of the intelligent road identification module. The problems of low efficiency, high cost, high subjectivity and the like in existing road inspection are solved, intelligent matching, order receiving and execution of road inspection tasks are realized by introducing a mode of combining an intelligent road inspection detection system and an internet platform, the inspection efficiency and accuracy are improved, and online management of a whole process is realized.

Description

Technical Field

[0001] The present application relates to the field of road inspection technology, and more specifically, to an intelligent road inspection method, system and storage medium. Background Art

[0002] With the acceleration of urbanization, the scale of road infrastructure continues to expand, and road inspection work has become increasingly arduous.

[0003] The current road inspection methods are all based on manual inspections. Traditional manual inspection methods have problems such as low efficiency, high cost, and strong subjectivity. It is difficult to meet the high requirements of modern cities for road inspections. And without a dedicated platform for management, it is difficult to efficiently distribute inspection tasks to matching users. It is difficult to understand the inspection process in a timely manner. Summary of the invention

[0004] The embodiments of the present application provide an intelligent road inspection method, system and storage medium to at least solve the technical problems of low efficiency, high cost and strong subjectivity existing in the road inspection of the prior art.

[0005] According to one aspect of an embodiment of the present application, there is provided an intelligent road inspection method, which is applied to an Internet platform, and includes:

[0006] Receive inspection task information issued by the first user and inspection service data of the second user;

[0007] Based on the pre-trained intelligent matching model, a matching calculation is performed on the inspection task information and the inspection service data to obtain a second user that matches the inspection task posted by the first user;

[0008] Sending the inspection task to the matched second user, so that the matched second user automatically identifies road damage information based on a road intelligent identification module;

[0009] Receive the inspection progress video, image, sensor data and road hazard identification information uploaded in real time during the inspection process of the road intelligent recognition module.

[0010] In one embodiment, based on a pre-trained intelligent matching model, matching calculation is performed on the inspection task information and the inspection service data to obtain a second user matching the inspection task posted by the first user, including:

[0011] Input the inspection task information and inspection service data into the pre-trained intelligent matching model; wherein the inspection task information includes inspection requirements, inspection location, and inspection completion time; the inspection service data includes user location, equipment capabilities, and historical evaluation;

[0012] The intelligent matching model calculates the geographic location matching degree based on the distance between the location of each second user and the inspection location; calculates the device capability matching degree based on the matching degree between the device capability of each second user and the inspection requirement; and calculates the historical evaluation matching degree based on the historical evaluation of each second user;

[0013] A weighted sum is performed based on the geographical location matching degree, the device capability matching degree, the historical evaluation matching degree, and the set weights to obtain a comprehensive matching degree of the second user;

[0014] The intelligent matching model sorts the comprehensive matching degrees from high to low, and outputs a preset number of second users at the front.

[0015] In one embodiment, it further includes:

[0016] After the second user confirms to accept the order, order information is generated; and a fee settlement interface and an after-sales service interface are provided; the order information includes the order number, order amount, payment status and payment time;

[0017] Receiving the road inspection report uploaded by the matched second user;

[0018] Receive the acceptance data uploaded by the first user.

[0019] In one implementation, after receiving the road inspection report uploaded by the second user, the method further includes:

[0020] Obtain road inspection reports for target roads within a preset time period;

[0021] Inputting the road inspection report into a preset evaluation model to obtain a road condition report;

[0022] predicting a road problem based on the road condition report and generating preventive measures corresponding to the road problem;

[0023] The road condition report, predicted road problems and corresponding preventive measures are sent to the first user.

[0024] In one embodiment, before receiving the inspection task information issued by the first user and the inspection service data of the second user, the method further includes:

[0025] Receiving registration information and authentication information of the first user and the second user;

[0026] Conduct qualification review on the registration information and certification information;

[0027] The permissions of the corresponding user identities are respectively opened for the first user and the second user information that have passed the review.

[0028] In one implementation, the second user automatically identifies road damage information based on the road intelligent identification module, including:

[0029] Acquire a road image acquired by an image acquisition device disposed on the inspection device;

[0030] The road intelligent recognition module recognizes the road image based on a preset road recognition model to obtain the type and location of road damage.

[0031] According to another aspect of the embodiment of the present application, there is provided an intelligent road inspection method, which is applied to a road intelligent identification module, comprising:

[0032] Acquire a road image acquired by an image acquisition device disposed on the inspection device;

[0033] The collected road images are input into the pre-trained road recognition model to obtain the type and location identification information of road damage;

[0034] Send road hazard type and location identification information to an Internet platform;

[0035] The inspection progress video, images, and sensor data generated in real time during the inspection process are sent to the Internet platform.

[0036] In one implementation, the road recognition model is obtained by training with a YOLOv8 neural network, and the road recognition model training method includes:

[0037] Collect road image data containing various types of road surface damage under different weather, lighting, and road surface material backgrounds;

[0038] Label the collected road image data to mark the types and locations of various road surface defects;

[0039] Preprocess the annotated road image data and divide it into training set and test set;

[0040] Training is performed based on the training set, test set data and YOLOv8 neural network, and model parameters are adjusted through a back propagation algorithm to obtain a trained road recognition model.

[0041] According to another aspect of the embodiment of the present application, there is provided an intelligent road inspection system, comprising:

[0042] An Internet platform is used to receive the inspection task information published by a first user and the inspection service data of a second user; input the inspection task information and the inspection service data into a pre-trained intelligent matching model to obtain a second user matching the inspection task; and send the inspection task to the matching second user;

[0043] The road intelligent identification module is used to automatically identify road damage information; send the type and location identification information of road damage to the Internet platform; and send the inspection progress video, image, and sensor data generated in real time during the inspection process to the Internet platform.

[0044] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the above-mentioned intelligent road inspection method when running.

[0045] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0046] A road inspection method provided in an embodiment of the present application introduces a mode combining a road intelligent identification module with an Internet platform. Through the road intelligent identification module, road disease information can be automatically identified, which greatly reduces manual intervention, improves the degree of automation of inspections, avoids subjective interference in manual inspections, improves the accuracy of identification, and reduces labor costs. Intelligent order acceptance, allocation, and execution of tasks are realized through the Internet platform, which reduces dependence on a large number of inspection personnel, reduces labor costs, and can efficiently assign inspection tasks to appropriate users. Users can publish inspection tasks, view inspection progress, road disease identification results, etc. on the Internet platform, realize online management of the entire process, and improve user experience.

[0047] Furthermore, the platform collects, organizes and analyzes inspection data to form a comprehensive road condition report, providing strong support for government decision-making, road maintenance and other work. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0049] Figure 1 is a flow chart of an intelligent road inspection method applied to an Internet platform according to an embodiment of the present application;

[0050] Figure 2 is a flow chart of an intelligent road inspection method applied to an intelligent road inspection detection module according to an embodiment of the present application;

[0051] Figure 3 is a flow chart of an intelligent road inspection method according to an embodiment of the present application;

[0052] Figure 4 is a diagram of a method for training and applying a road recognition model according to an embodiment of the present application;

[0053] Figure 5 is a schematic diagram of an intelligent matching method according to an embodiment of the present application;

[0054] Figure 6 It is an interactive schematic diagram of an intelligent road inspection method according to an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0056] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0057] The following is combined with Figure 1 The intelligent road inspection method of the embodiment of the present application is described in detail. Figure 1 As shown, when applied to an Internet platform, the method mainly includes the following steps:

[0058] S101 receives inspection task information issued by a first user and inspection service data of a second user.

[0059] The first user refers to the demander who publishes the road inspection task, including government departments, road maintenance units, individuals, etc. The second user is the service provider who performs the inspection task.

[0060] In one embodiment, before receiving the inspection task information published by the first user and the inspection service data of the second user, it also includes: receiving the registration information and authentication information of the first user and the second user; performing a qualification review on the registration information and the authentication information; and opening up the corresponding user identity permissions for the first user and the second user information that have passed the review.

[0061] Users register an account through the Internet platform and submit relevant authentication information, such as corporate qualifications, personal ID card, etc. The platform reviews the user information, confirms their identity and qualifications, and grants corresponding permissions. Users can be the first user or the second user.

[0062] Furthermore, the inspection service data of the second user is pre-registered on the Internet platform, and its inspection service data such as service capabilities and service scope can be updated in real time. When the Internet platform receives the inspection service needs of the first user, it can match the appropriate target second user according to the real-time inspection service provision on the platform.

[0063] Furthermore, the first user logs in to the Internet platform to publish inspection task information, which includes inspection requirements, inspection locations, inspection completion time, inspection task urgency, etc. The platform stores the task information in a database and generates a task number, which can be stored in a message queue and sorted according to the release time and urgency of the inspection task.

[0064] Furthermore, the inspection service data of the second user is obtained, including user location, equipment capability, historical evaluation, and service price and other information.

[0065] S102 performs matching calculation on the inspection task information and the inspection service data based on the pre-trained intelligent matching model to obtain a second user that matches the inspection task posted by the first user.

[0066] In one embodiment, the inspection task information and inspection service data are input into the pre-trained intelligent matching model. The inspection task information includes inspection requirements, inspection location, and inspection completion time; the inspection service data includes user location, equipment capabilities (such as the types of services that can be provided, equipment types), and historical evaluations (customer feedback, service completion quality, etc.).

[0067] Furthermore, the collected data is cleaned to remove duplicate, erroneous or inconsistent data to ensure data accuracy and consistency. At the same time, the data is standardized for subsequent calculation and analysis.

[0068] The intelligent matching model calculates the geographic location matching degree based on the distance between the location of each second user and the inspection location; calculates the equipment capability matching degree based on the matching degree between the equipment capability of each second user and the inspection requirements; calculates the historical evaluation matching degree based on the historical evaluation of each second user; and obtains the comprehensive matching degree of the second user by weighted summation based on the geographic location matching degree, equipment capability matching degree, historical evaluation matching degree and set weights.

[0069] Specifically, the geographic location matching degree is usually calculated based on the actual distance or time cost between the user and the task segment, and is represented by the inverse of the distance or the decay function of the distance.

[0070] For example, the geographic location matching degree is calculated by the following formula: Wloc=1+α·d1;

[0071] Where d1 is the straight-line distance between the user and the task segment, or the equivalent distance adjusted according to traffic conditions. α is an adjustment factor used to control the degree of influence of distance on geographic location matching. When d1 approaches 0, Wloc approaches 1, indicating that the geographic locations are very close; as d1 increases, Wloc gradually decreases.

[0072] Calculate the device capability match, which reflects the ability of the user's device to meet specific needs. It is evaluated based on the performance parameters of the device (such as processor speed, memory size, screen resolution, etc.). Example formula (assuming there is a comprehensive score Sdev).

[0073] Wdev=Smax-SminSdev-Smin;

[0074] Among them, Sdev is the comprehensive score of the user's device, Smin and Smax are the minimum and maximum scores of all devices respectively. This formula normalizes the device capability score to between 0 and 1.

[0075] Calculate the historical evaluation matching degree, which is based on the evaluation of the user's past completed tasks. It is achieved by taking the average rating and the weighted average of the most recent ratings.

[0076] Example formula (using average rating Ravg):

[0077] Whist = Rmax - RminRavg - Rmin;

[0078] Among them, Ravg is the average score of the user's past completed tasks, Rmin and Rmax are the minimum and maximum values ​​in the scoring system respectively. This formula also normalizes the score to between 0 and 1.

[0079] Calculate the comprehensive matching degree, the comprehensive matching score Stotal example:

[0080] Stotal=β1·Wloc+β2·Wdev+β3·Whist;

[0081] Among them, β1, β2, and β3 are weight coefficients set according to business requirements, and β1+β2+β3=1. These coefficients reflect the relative importance of different factors in the matching process. Optionally, price factors and response time factors can also be added.

[0082] The intelligent matching model sorts the users from high to low according to the comprehensive matching degree, and outputs a preset number of second users who are ranked in front. Based on the sorting results, the platform can automatically select the user with the highest comprehensive matching degree for matching, or select the top few users as candidates according to actual needs.

[0083] like Figure 5 The figure shows a schematic diagram of an intelligent matching method. First, data collection and preprocessing, including collecting user data and inspection demand data. Further, data cleaning and standardization are performed.

[0084] Further, the weight of the matching factor is set, the weight of the geographical location, the weight of the device capability, and the weight of the historical evaluation are set, and the geographical location matching degree, the device capability matching degree, and the historical evaluation matching degree are calculated. Other factors may also be included, and the embodiments of the present application are not specifically limited. The comprehensive matching degree is obtained according to the set factors. The results are sorted and intelligently selected according to the comprehensive matching degree of each user.

[0085] Based on the user's geographical location, equipment capabilities, historical evaluation and other factors, as well as the inspection needs of the maintenance department, the system intelligently matches and assigns tasks to the most suitable users to ensure that the tasks can be completed efficiently and accurately. This improves the efficiency and success rate of task allocation and order acceptance, and improves customer experience.

[0086] S103 sends the inspection task to the matched second user, so that the matched second user can automatically identify the road damage information based on the road intelligent recognition module.

[0087] In one embodiment, the successfully matched task is assigned to the corresponding second user, and the second user is notified via SMS, email, or APP push, etc. If the user confirms the order, a prompt message indicating that the order is successfully accepted can be returned.

[0088] After the second user successfully accepts the order, he can operate the AI ​​intelligent road patrol detection system to patrol the designated area and automatically identify road damage information. The AI ​​intelligent road patrol detection system includes a road intelligent recognition module, which stores a pre-trained road recognition model based on the YOLOv8 neural network; obtains road images from an image acquisition device installed in the patrol device; and recognizes the collected road images based on the pre-trained road recognition model to obtain the type and location of road damage.

[0089] In the implementation manner of the present application, the inspection device can be an inspection vehicle or a drone; the image acquisition device can be a camera, a lidar, an infrared sensor, and the camera can be a vehicle-mounted camera or a handheld mobile camera device, such as a camera mounted on a mobile phone.

[0090] This application uses artificial intelligence technology for road inspections, which can automatically identify road conditions, greatly reduce manual intervention, and improve the degree of automation of inspections. Based on the YOLOv8 pre-trained model, it is specifically used to automatically identify road disease characteristics. Through this process, a large amount of road disease data is prepared, including transverse cracks, longitudinal cracks, block cracks, cracks, potholes, looseness, road subsidence and other road disease feature recognition training and detection. Through deep pre-training of these data, the AI ​​algorithm can achieve high-precision and high-stability recognition capabilities, can more accurately identify road conditions, and reduce human errors.

[0091] S104 receives the inspection progress video, image, sensor data and road hazard identification information uploaded in real time by the intelligent road patrol detection module during the inspection process.

[0092] During the inspection process, the road intelligent identification module will feed back the inspection results to the Internet platform in real time, and users can check the inspection progress and results at any time. Specifically, the inspection location, images, videos, sensor data collected during the inspection, etc. are sent to the Internet platform, and the road disease type and disease location data are sent to the Internet platform. At the same time, the platform provides real-time monitoring functions to ensure the smooth progress of the inspection work.

[0093] Optionally, the first user and the second user can use the chat interface provided by the Internet platform to discuss problems in the inspection process, thereby improving the timeliness of problem solving.

[0094] In one embodiment, the method further includes: receiving a road inspection report uploaded by a second user, the road inspection report including data such as the inspection road ID, inspection time, road problems found, and the location of the problem road. The first user can obtain the road inspection report from the platform, inspect the completion of the road inspection task, and upload the inspection data to the Internet platform, and the Internet platform receives the inspection data uploaded by the first user.

[0095] The Internet platform provided by this application can conduct online transactions. After the second user confirms to accept the order, the order information is generated; and a fee settlement interface and an after-sales service interface are provided; the order information includes the order number, order amount, payment status, and payment time. The user's payment voucher, payment channel, payment time, etc. are also queried. A fee settlement interface and an after-sales service interface are provided. If there is a refund or complaint, the relevant information and processing results must be recorded. The Internet platform of this application can provide online management of the entire process, which greatly improves the user experience, service efficiency, and convenience.

[0096] In one embodiment, after receiving the road inspection report uploaded by the second user, it also includes: obtaining a road inspection report for the target road within a preset time period; inputting the road inspection report into a preset evaluation model to obtain a road condition report; predicting road problems based on the road condition report, and generating preventive measures corresponding to the road problems; sending the road condition report, predicted road problems, and corresponding preventive measures to the first user. This may be a road maintenance department or a government agency. Such information transmission ensures that relevant decision makers can obtain necessary information in a timely manner in order to take appropriate actions. This process not only improves the efficiency of road inspections, but also enhances the foresight of road maintenance through predictive analysis, which helps to reduce road accidents and maintenance costs. In this way, the AI ​​intelligent road patrol detection system and the Internet order-taking model can provide a more intelligent and automated solution for road maintenance.

[0097] In one embodiment, road inspection reports for the past three months can be obtained, and road disease information detected can be obtained based on multiple road inspection reports. These data are input into the evaluation model. The evaluation model evaluates the road surface condition, the integrity of road traffic facilities, the road drainage condition, the roadbed condition, including the stability of the roadbed and the performance of the road surface materials. The evaluation model integrates the road surface condition evaluation method, including calculating the number of transverse cracks, longitudinal cracks, block cracks, cracks, potholes, looseness, road surface subsidence and other diseases, and determining the corresponding road surface condition information (damaged, good, slightly damaged) according to the number interval, and integrates the road traffic facility evaluation method to check whether traffic signs, markings, signal lights, guardrails, etc. are clear and intact, obtain the number of damaged traffic facilities, and determine the corresponding traffic facility condition information "a little damaged, a lot of damaged, no damaged" according to the number interval.

[0098] Predict road problems based on road condition assessment information, and build prediction models using machine learning and deep learning techniques, such as convolutional neural networks (CNN) or recurrent neural networks (RNN). The model is able to learn historical data and patterns of road conditions to predict possible problems in the future. A method for urban road condition perception and short-term prediction based on multi-source data fusion can be built. Model training and verification: Use the collected data to train the model and test the accuracy of the model through a validation set. The model can be verified by comparing the results predicted by the model with the actual road problems that have occurred. Apply the trained model to real-time data to predict road problems. Query the expert knowledge base based on the predicted problems and give corresponding preventive suggestions.

[0099] like Figure 2 As shown, according to another aspect of the present application, an intelligent road inspection method is provided, which is applied to a road intelligent identification module, comprising:

[0100] S201 obtains a road image acquired by an image acquisition device provided in the inspection device;

[0101] S202: inputting the collected road image into a pre-trained road recognition model to obtain the type and location recognition information of road damage;

[0102] S203 sends the type and location identification information of the road damage to the Internet platform;

[0103] S204 sends the inspection progress video, images, and sensor data generated in real time during the inspection process to the Internet platform.

[0104] The road intelligent recognition module corresponding to the technical solution of the present application has functions such as image acquisition, transmission, image recognition, data processing and analysis, and can be a minimum product unit capable of intelligently recognizing road image information.

[0105] The AI ​​intelligent road patrol detection system of the present application has hardware components including patrol devices, image acquisition equipment, sensors (such as laser radar, infrared sensor, etc.), GPS positioning system, etc., wherein the patrol device can be a patrol vehicle or a drone; the image acquisition equipment can be a camera, laser radar, infrared sensor, the camera can be a car camera, or a handheld mobile camera device, such as a mobile phone. It also includes a road intelligent recognition module for road disease recognition.

[0106] In one embodiment, the road recognition model training method includes: collecting road image data containing various types of road surface diseases under different weather, lighting, and road surface material backgrounds. Collecting image and video data containing various road surface diseases (such as transverse cracks, longitudinal cracks, block cracks, cracks, potholes, looseness, road surface subsidence, etc.) from multiple channels. Ensure that the data covers multiple scenarios such as different weather conditions, lighting intensity, road surface materials, etc. to improve the generalization ability of the model.

[0107] The collected road image data is labeled to mark the types and locations of various road surface defects. During the labeling process, it is necessary to ensure the accuracy and consistency of the labeling so that the model can learn effective features.

[0108] Preprocess the annotated road image data and divide them into training sets and test sets. Perform preprocessing operations such as denoising, grayscale, and contrast enhancement on the annotated images to improve image quality and reduce the impact of noise on model training. Crop and scale the images to a uniform size as needed to meet the model input requirements.

[0109] YOLOv8 is selected as the pre-trained model architecture because of its efficient and accurate target detection capabilities. According to the specific needs of pavement disease detection, the YOLOv8 model is adjusted and optimized as necessary.

[0110] Based on the training set, test set data and YOLOv8 neural network, the model parameters are adjusted through the back propagation algorithm to obtain a trained road recognition model. During the training process, the model is regularly evaluated using the validation set data to promptly detect and correct problems such as overfitting. The model is tuned based on the results of the validation set and test set. This may include adjusting the learning rate, optimizer, regularization strategy, etc. Cross-validation and other techniques are used to further improve the generalization ability of the model.

[0111] Next, we enter the model evaluation and testing phase, using the test set data to evaluate the performance of the trained model, including indicators such as accuracy, recall, and F1 score. We evaluate the performance of the model in different scenarios, such as different weather conditions and light intensity.

[0112] Test the model in actual road detection scenarios to verify its effectiveness and reliability in practical applications. Collect feedback and further optimize and improve the model.

[0113] Deploy the trained model to the actual road surface disease detection system. Use the deployed model to automatically detect the road surface images collected in real time and identify various road surface diseases. Output the detection results in the form of image annotation and report to provide timely and accurate disease information for road maintenance departments.

[0114] Regularly collect new pavement disease data, update and retrain the model to adapt to new detection needs and environmental changes. Regularly maintain and optimize the model to ensure its long-term stable operation.

[0115] like Figure 4 As shown, the model training and application methods include:

[0116] S401 training data collection and labeling; S402 data preprocessing; S403 selection of neural network model architecture; S404 data set division; S405 training model based on divided data set; S406 model tuning; S407 model evaluation and testing; S408 model deployment and application; S409 model maintenance and update.

[0117] According to another aspect of the embodiment of the present application, there is also provided an intelligent road inspection system, including:

[0118] The Internet platform is used to receive the inspection task information published by the first user and the inspection service data of the second user; input the inspection task information and the inspection service data into the pre-trained intelligent matching model to obtain the second user matching the inspection task; and send the inspection task to the matching second user;

[0119] The road intelligent identification module is used to automatically identify road damage information; send the type and location identification information of road damage to the Internet platform; and send the inspection progress video, images, and sensor data generated in real time during the inspection process to the Internet platform.

[0120] like Figure 3 As shown, the implementation steps based on the intelligent road inspection system include:

[0121] S301 User registration and authentication: Users register an account through the Internet platform and submit relevant authentication information (such as corporate qualifications, personal ID card, etc.). The platform reviews the user information, confirms their identity and qualifications, and grants corresponding permissions.

[0122] S302 Release road inspection tasks; the user logs in to the platform and fills in the relevant information of the inspection task, including the inspection area, inspection requirements, completion time, etc. The platform stores the task information in the database and generates a task number.

[0123] S303 Intelligent matching and task allocation: The platform performs intelligent matching based on factors such as the service provider’s geographical location, equipment capabilities, historical evaluation, and the inspection needs of the maintenance department. Successfully matched tasks are assigned to the corresponding users, and the users are notified via SMS, email, or APP push.

[0124] S304 Task execution: After receiving the task, the user uses the AI ​​intelligent road patrol detection system to go to the designated area for inspection. The system collects road images through high-definition cameras such as CCD and CMOS, and uses deep learning models for real-time recognition and analysis. The recognition results (such as road diseases, road abnormalities, etc.) are uploaded to the Internet platform in real time.

[0125] S305 Real-time feedback and monitoring: Users can view the inspection progress and results in real time through the platform, including inspection locations, identified problems, etc. The platform provides real-time monitoring functions to ensure the smooth progress of inspection work.

[0126] S306 Task acceptance and settlement. After the inspection is completed, the user submits the inspection report to the platform. The maintenance department accepts the inspection report and completes the task after confirming that it is correct. The platform automatically settles the fees based on the completion of the task and generates a settlement report.

[0127] like Figure 6 As shown in the figure, the user interaction methods based on the intelligent road inspection system include:

[0128] First, the user logs in to the Internet platform for registration and authentication. The Internet platform stores the user information in the database. The Internet platform confirms the user's identity and authority. The user logs in to the Internet platform to publish the task. The Internet platform stores the task information in the database. The Internet platform generates a task number and performs intelligent matching and task assignment. Send the task assignment notification to the user. The user performs the task based on the AI ​​intelligent road patrol detection system. And upload the recognition results to the Internet platform. The Internet platform stores the inspection results in the database. The user feeds back the inspection process to the Internet platform in real time. Task acceptance and settlement are carried out based on the Internet platform. The Internet platform sends the inspection report to the maintenance department. The Internet platform automatically settles the fees based on the acceptance report sent by the maintenance department, generates a settlement report and stores it in the database. Realize online management of the entire process.

[0129] Through the above implementation methods, the Internet order-taking mode based on the intelligent road patrol detection system can realize the intelligent order-taking, allocation and execution of road patrol tasks. This mode not only improves the efficiency and accuracy of patrol inspections, but also reduces labor costs and time costs, providing strong support for government decision-making, road maintenance and other work. At the same time, this model is also scalable and customizable, and can be expanded and customized according to different needs.

[0130] According to another aspect of the embodiments of the present application, a computer-readable storage medium corresponding to the intelligent road inspection method provided in the aforementioned embodiments is also provided, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it will execute the intelligent road inspection method provided in any of the aforementioned embodiments.

[0131] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0132] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the intelligent road inspection method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0133] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the attached claims.

Claims

1. An intelligent road inspection method, applied to an Internet platform, characterized in that: include: Receive inspection task information issued by the first user and inspection service data of the second user; Based on the pre-trained intelligent matching model, a matching calculation is performed on the inspection task information and the inspection service data to obtain a second user that matches the inspection task posted by the first user; Sending the inspection task to the matched second user, so that the matched second user automatically identifies road damage information based on a road intelligent identification module; Receive the inspection progress video, image, sensor data and road hazard identification information uploaded in real time during the inspection process of the road intelligent recognition module.

2. The method according to claim 1, characterized in that Based on the pre-trained intelligent matching model, matching calculation is performed on the inspection task information and the inspection service data to obtain a second user matching the inspection task issued by the first user, including: Input the inspection task information and inspection service data into the pre-trained intelligent matching model; wherein the inspection task information includes inspection requirements, inspection location, and inspection completion time; the inspection service data includes user location, equipment capabilities, and historical evaluation; The intelligent matching model calculates the geographic location matching degree based on the distance between the location of each second user and the inspection location; calculates the device capability matching degree based on the matching degree between the device capability of each second user and the inspection requirement; and calculates the historical evaluation matching degree based on the historical evaluation of each second user; A weighted sum is performed based on the geographical location matching degree, the device capability matching degree, the historical evaluation matching degree, and the set weights to obtain a comprehensive matching degree of the second user; The intelligent matching model sorts the comprehensive matching degrees from high to low, and outputs a preset number of second users at the front.

3. The method according to claim 1, characterized in that Also includes: After the second user confirms to accept the order, order information is generated; And provide fee settlement interface and after-sales service interface; The order information includes order number, order amount, payment status and payment time; receiving the road inspection report uploaded by the matched second user; Receive the acceptance data uploaded by the first user.

4. The method according to claim 3, characterized in that After receiving the road inspection report uploaded by the second user, the method further includes: Obtain road inspection reports for target roads within a preset time period; Inputting the road inspection report into a preset evaluation model to obtain a road condition report; predicting a road problem based on the road condition report and generating preventive measures corresponding to the road problem; The road condition report, predicted road problems and corresponding preventive measures are sent to the first user.

5. The method according to claim 1, characterized in that Before receiving the inspection task information published by the first user and the inspection service data of the second user, the method further includes: Receiving registration information and authentication information of the first user and the second user; Conduct qualification review on the registration information and certification information; The permissions of the corresponding user identities are respectively opened for the first user and the second user information that have passed the review.

6. The method according to claim 1, characterized in that The second user automatically identifies road damage information based on the road intelligent identification module, including: Acquire a road image acquired by an image acquisition device disposed on the inspection device; The road intelligent recognition module recognizes the road image based on a preset road recognition model to obtain the type and location of road damage.

7. An intelligent road inspection method, applied to a road intelligent identification module, characterized in that: include: Acquire a road image acquired by an image acquisition device disposed on the inspection device; The collected road images are input into the pre-trained road recognition model to obtain the type and location identification information of road damage; Send road hazard type and location identification information to an Internet platform; The inspection progress video, images, and sensor data generated in real time during the inspection process are sent to the Internet platform.

8. The method according to claim 7, characterized in that The road recognition model is obtained by training with a YOLOv8 neural network, and the road recognition model training method includes: Collect road image data containing various types of road surface damage under different weather, lighting, and road surface material backgrounds; Label the collected road image data to mark the types and locations of various road surface defects; Preprocess the annotated road image data and divide it into training set and test set; Training is performed based on the training set, test set data and YOLOv8 neural network, and model parameters are adjusted through a back propagation algorithm to obtain a trained road recognition model.

9. An intelligent road inspection system, characterized in that: include: An Internet platform for receiving inspection task information published by a first user and inspection service data of a second user; Input the inspection task information and inspection service data into a pre-trained intelligent matching model to obtain a second user matching the inspection task; and send the inspection task to the matching second user; The road intelligent identification module is used to automatically identify road damage information; send the type and location identification information of road damage to the Internet platform; and send the inspection progress video, image, and sensor data generated in real time during the inspection process to the Internet platform.

10. A computer-readable medium, characterized in that Computer-readable instructions are stored thereon, and the computer-readable instructions are executed by a processor to implement an intelligent road inspection method as described in any one of claims 1 to 8.

Images