Urban road intelligent inspection method and system based on automatic driving vehicle

Abstract

The invention discloses an urban road intelligent inspection method and system based on an automatic driving vehicle. The method comprises the following steps: S1, acquiring position information of the automatic driving vehicle through a positioning module loaded at a vehicle end; s2, sensing the surrounding environment of the vehicle through a camera module loaded on the vehicle end so as to construct image information; s3, target detection is carried out based on the constructed image information through a deep learning algorithm, whether road abnormity exists or not is judged, if yes, the step S4 is carried out, and if not, the step S1 is carried out; s4, the vehicle end organizes the vehicle information and uploads the vehicle information to the cloud platform data center; and S5, after the cloud platform data center receives the corresponding information, carrying out data induction and arrangement, data storage, construction of a maintenance task and tracking. According to the technical scheme, a brand-new, more intelligent and efficient infrastructure maintenance mode is provided for urban managers, and promotion of urban traffic and living environment is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of urban intelligent inspection technology, and in particular to an urban road intelligent inspection method and system based on an autonomous driving vehicle. Background Art

[0002] Urban road infrastructure is the most basic construction for urban development and operation, and constitutes an indispensable part of urban transportation and living environment. This covers a number of key elements, including road surfaces, green belts, street lights, traffic signs and markings, bus stops and public facilities, manhole covers and drainage wells, etc. These facilities together build the city's transportation network and public service system, providing necessary support for the lives and travel of urban residents. These infrastructures are under tremendous pressure during urban operations. The natural environment (heavy rain, storms), traffic pressure, man-made damage, aging and lack of maintenance and other problems will lead to certain safety hazards in the infrastructure, which requires frequent inspections and maintenance.

[0003] At present, infrastructure maintenance in urban roads faces many difficulties and challenges, which directly affect the smoothness of urban traffic, the quality of life of residents, and the completeness and long-term sustainability of infrastructure.

[0004] 1. Inefficient manual inspection: Traditional infrastructure maintenance mainly relies on manual inspection, which is inefficient. Urban roads are crisscrossed, and the huge infrastructure network is difficult to monitor comprehensively with limited manpower and time. This leads to some potential problems that cannot be discovered and handled in time, which in turn evolves into more serious hidden dangers.

[0005] 2. Missing small but dangerous problems: Manual inspections can easily miss some small but potentially dangerous problems, such as displacement of manhole covers, cracks in the road surface, dim street lights, and plants in green belts blocking the road, which may not be easily noticed by human eyes.

[0006] 3. High maintenance costs: Traditional maintenance methods require a lot of manpower, material and financial resources, with a long maintenance cycle and high maintenance costs. This not only limits the timely maintenance of infrastructure, but also increases the burden on city managers.

[0007] 4. Inability to provide full coverage: Due to the large and complex urban road network, traditional maintenance methods cannot achieve full coverage. Infrastructure in some remote areas or areas with little traffic is easily overlooked, resulting in unfair resource allocation and service coverage.

[0008] State Grid Shandong Electric Power Company Jinan Power Supply Company disclosed a smart city IoT manhole cover data control system and method based on edge computing on October 10, 2023 (see patent CN202311126159.3 for details). It relates to the field of urban data control technology, specifically a smart city IoT manhole cover data / control system and method based on edge computing. The system includes an edge analysis feedback module, which collects the status data information of the electronic manhole cover in real time through the built-in sensor of the electronic manhole cover, and analyzes the collected status data information through the edge computing unit in the electronic manhole cover, and judges the operating status of the electronic manhole cover in combination with the control command data sent by the IoT control terminal for the most recent time received by the electronic manhole cover, and feeds back the status data information corresponding to the electronic manhole cover with abnormal operating status to the IoT control terminal.

[0009] China University of Geosciences (Wuhan) disclosed a drone urban inspection system and method on November 11, 2022 (see patent CN202211409069.0 for details), including: a storage management module, used to upload and save the target data set to the underlying storage unit, and the underlying storage unit saves the target detection model and the drone flight path; a distributed cache module, used to cache the target data set to the distributed cache unit; a distributed training module, used to read the target data set and perform distributed training on the target detection model; a path planning module, used to calculate and plan the drone flight path, and control the drone to cruise along the drone flight path; a target detection module, used to detect the detection target using the target detection model.

[0010] Suzhou Futai Information Technology Co., Ltd. disclosed a city intelligent patrol vehicle system and patrol method based on visual AI technology on March 11, 2022 (see patent CN202210243060.0 for details). Its patrol system includes a vehicle-mounted host and a data acquisition module connected to the vehicle-mounted host. The vehicle-mounted host is wirelessly connected to the central server to parse and process the instructions received from the central server and send alarm information to the central server. The data acquisition module includes a visual sensor, a vehicle-mounted camera and a vehicle-mounted GNSS. The vehicle-mounted host is equipped with an image preprocessing module, a target detection module and a target tracking module. The target detection module uses the YOLOv5 algorithm. The patrol vehicle system can realize real-time detection of illegal parking, mobile vendors, exposed garbage and other violations.

[0011] Shanghai Piaoshi Network Co., Ltd. disclosed a smart city road debris detection device on September 9, 2022 (see patent CN202220376863.9 for details). Its main shell is mounted on the vertical pole of the luggage rack on the top of the vehicle through an installation mechanism. It is powered by a battery or solar panel. The shell is provided with a high-definition camera facing the road surface. An image processor, a single-chip microcomputer and a power supply battery are provided in the shell. The image processor is connected to the single-chip microcomputer. The single-chip microcomputer is provided with an image recognition unit and an image storage unit. The single-chip microcomputer is connected to an alarm, and the power supply battery is electrically connected to the high-definition camera, image processor, single-chip microcomputer and alarm respectively.

[0012] Judging from the existing publicly available patent technologies, the entire detection process has poor scalability and adaptability, and there is no timely follow-up maintenance method, and lacks data statistics and analysis.

[0013] 1. Poor scalability: As described in the smart city IoT manhole cover data management and control system and method based on edge computing, it is necessary to install terminal intelligent devices on each manhole cover, which consumes a lot of manpower and financial resources.

[0014] 2. The urban inspection method based on drones is greatly limited by the flight endurance of drones. It may not be possible to complete the identification of the coverage area while ensuring accuracy in a single flight operation.

[0015] 3. The single-camera vision method is limited by the fixed installation method, with a limited field of view and a small detection range. It is also affected by factors such as lighting and weather, and may perform poorly in some cases.

[0016] 4. There is no effective maintenance and follow-up tracking of the detected abnormal facilities, and there is a lack of accumulation and analysis of data. Summary of the invention

[0017] In order to solve the above technical problems, the present invention proposes an urban road intelligent inspection method based on an autonomous driving vehicle, which comprises the following steps:

[0018] S1: Obtain the location information of the autonomous driving vehicle through the positioning module installed on the vehicle;

[0019] S2: The camera module mounted on the vehicle side is used to sense the surrounding environment of the vehicle to construct image information;

[0020] S3: Using a deep learning algorithm, based on the constructed image information, perform target detection and determine whether there is a road abnormality. If yes, proceed to step S4, otherwise proceed to step S1;

[0021] S4: The vehicle side organizes the vehicle information and uploads the vehicle information to the cloud platform data center;

[0022] S5: After receiving the corresponding information, the cloud platform data center summarizes and organizes the data, stores the data in the database, builds maintenance tasks and tracks them.

[0023] Furthermore, in step S1, the positioning module includes GNSS and IMU sensor devices.

[0024] Furthermore, step S1 includes: calculating the acceleration and angular velocity of the vehicle in three spatial directions through the IMU to provide the vehicle's attitude, linear acceleration and angular velocity information; calculating the absolute position information of the vehicle through the GNSS; and calculating the high-frequency precise position information of the vehicle by fusing the high-frequency three-axis attitude angle and acceleration information provided by the IMU and the absolute position information provided by the GNSS.

[0025] Furthermore, in step S3, the daily collected data is labeled and anomalies are marked. The abnormal targets and target types in the pictures need to be marked. The labeled image data is used to train a vision-based target detection deep learning model, and a target detection unit is constructed for target detection during the driving of the autonomous driving vehicle to mark the target object and the abnormal frame; the depth of the abnormal frame of the output image in the vehicle body coordinate system is output, and a multi-view BEV network is constructed, that is, by projecting the target object mark and the abnormal frame into the vehicle body coordinate system.

[0026] Furthermore, in step S3, the abnormal road conditions include: displacement of manhole covers, foreign objects on the road surface, cracks on the road surface, and damage to traffic markings and signs.

[0027] Furthermore, in step S4, the self-vehicle information includes the current position information of the autonomous driving vehicle, the target detection result of the target detection model, and the image information; and the vehicle end uploads the self-vehicle information to the cloud platform data center through the communication module.

[0028] Furthermore, in step S5, after receiving the uploaded information, the cloud platform data center classifies and organizes the data according to the target detection results, determines the risk level, and automatically constructs maintenance tasks and assigns them to different maintenance personnel.

[0029] Furthermore, in step S4, the cloud platform filters and analyzes the confidence of road anomalies at specific locations based on multiple reports from multiple vehicles and geographical locations. When the confidence is high, the risk level is determined based on the category and geographical location. When the confidence is low, the information is transferred to manual confirmation.

[0030] Furthermore, the method also includes the following steps: S6: After the maintenance personnel perform on-site maintenance, the maintenance tasks of the cloud platform data center are updated according to the actual maintenance situation.

[0031] The technical solution of the present invention also provides an urban road intelligent inspection system based on an autonomous driving vehicle, which includes an autonomous driving vehicle and a cloud platform data center, wherein the autonomous driving vehicle is equipped with a positioning module, a camera module and a communication module, wherein the system performs urban road intelligent inspection according to the method described above.

[0032] Compared with the existing technology, the implementation of the technical solution of the present invention will provide city managers with a new, more intelligent and efficient way of infrastructure maintenance, which is conducive to promoting the improvement of urban transportation and living environment. It has the following advantages:

[0033] 1. Efficient inspection: The introduction of autonomous vehicles has greatly improved inspection efficiency and achieved high-frequency coverage of every corner of the city.

[0034] 2. Comprehensive monitoring: Through multi-sensing modules, all types of infrastructure can be comprehensively monitored to ensure that every detail is fully paid attention to.

[0035] 3. Instant reporting: Using big data analysis and cloud monitoring platforms, problems can be discovered, reported, and resolved at an early stage, improving the timeliness of maintenance.

[0036] 4. Convenient maintenance: Establish an automated management cloud platform to fully record the safety status of various facilities, assign maintenance tasks with one click, and comprehensively track problems and solutions.

[0037] 5. Reduce costs: Automated inspections reduce labor costs and improve the economic benefits of urban infrastructure maintenance.

[0038] 6. Improve safety: Through real-time monitoring and timely reporting, potential safety hazards can be discovered and resolved earlier, and the overall safety level of urban roads can be comprehensively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 It is a structural schematic diagram of the urban road intelligent inspection system based on the autonomous driving vehicle of the present invention;

[0041] Figure 2 It is a flow chart of the intelligent inspection method of urban roads based on the autonomous driving vehicle of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] As mentioned in the background technology section, the maintenance methods of traditional urban road infrastructure face many difficulties, and more efficient and intelligent inspection and maintenance methods need to be introduced to better cope with the challenges in urban development. The present invention introduces intelligent inspection technology, using an autonomous driving vehicle as a carrier, whose operation route is relatively fixed and equipped with sensors such as high-precision cameras, to achieve intelligent monitoring of urban road infrastructure and conduct comprehensive tracking and tracing of various types of equipment through a management cloud platform.

[0044] Urban public services such as sanitation, logistics, public transportation, and security patrols, as important components of urban infrastructure, have always been subject to traditional, manpower-driven operating methods. In recent years, with the continuous evolution of autonomous driving technology, autonomous vehicles have emerged as an innovative solution. This technology integrates advanced perception systems, real-time data analysis, and intelligent path planning to bring more efficient and sustainable solutions to urban public services. In the operation tasks of autonomous vehicles, multi-sensor collaboration plays a key role, providing a solid foundation for their safe and efficient operation in complex urban environments. These multi-sensor systems have the ability to work together to perceive, interpret, and adapt to the surrounding environment in real time to ensure that the vehicle can effectively perform various urban public service tasks.

[0045] Intelligent data collection plays an important role in the practical application of autonomous driving. Vehicles use sensors such as cameras and GNSS to perceive the vehicle's position and surrounding geographical environment in real time, providing accurate information for autonomous driving planning and control. The collaboration between these sensors is the key to the success of autonomous driving vehicles. They can share perception data in real time and integrate the information of each sensor into a comprehensive environmental model. The model's understanding of the surrounding environment is the basis for autonomous driving decisions. It allows vehicles to analyze, evaluate and predict road conditions in real time to develop appropriate action plans. In addition, multi-sensor collaboration also enhances the decision-making ability of autonomous driving vehicles. Vehicles can autonomously choose appropriate driving routes, speeds and actions based on real-time feedback from sensor data to cope with different tasks and scenarios.

[0046] The present invention utilizes various high-performance sensors carried by autonomous driving vehicles to detect damage to urban infrastructure such as road surfaces, green belts, street lights, traffic signs and markings, bus stops and public facilities, manhole covers and drainage wells, and uploads abnormal data to the cloud platform. At the same time, the cloud platform constructs maintenance tasks and assigns maintenance personnel to respond quickly.

[0047] See also Figure 1 , which shows the hardware and facility composition of each link of the technical solution for intelligent urban road inspection of autonomous driving vehicles of the present invention.

[0048] Among them, I MU: Inertial Measurement Unit, which provides high-frequency multi-directional acceleration information for autonomous driving vehicles;

[0049] GNSS: provides accurate absolute position information for vehicles;

[0050] Camera: Peripheral camera and surround camera provide information about the surrounding environment along the vehicle's route;

[0051] Communication module: Real-time monitoring data of vehicles is uploaded to the cloud platform through the communication module;

[0052] Cloud platform data center: stores data uploaded by autonomous vehicles, points out equipment that may have hidden dangers, reminds relevant personnel to perform maintenance, and records damage to road facilities and the repair process.

[0053] See also Figure 2 During the operation of the autonomous driving vehicle, the function of intelligent inspection of urban roads is realized through the following process:

[0054] S1: The IMU calculates the three-axis attitude angle and acceleration of the autonomous driving vehicle, provides information such as the vehicle's attitude (including direction, tilt and rotation, etc.), linear acceleration and angular velocity, and combines the GNSS absolute position information to provide the vehicle with accurate and high-frequency self-positioning information;

[0055] S2: construct image information through the information of the camera module;

[0056] S3: Use deep learning algorithms to detect targets in image data and analyze whether there are abnormal targets in the lane lines, such as the protruding parts of tree branches in green belts, foreign objects on the road surface, cracks on the road surface, displacement of manhole covers, broken street lights, and broken traffic signs and markings;

[0057] S4: When damage or foreign objects are detected, the vehicle side organizes the vehicle information and organizes the positioning, target detection results, target annotation, and image information into data packets;

[0058] S5: Upload the data packet to the cloud platform data center through the communication module;

[0059] S6: After receiving the uploaded information, the cloud platform classifies and organizes the data according to different target detection results, determines the risk level, and automatically constructs maintenance tasks and assigns them to different maintenance personnel;

[0060] S7: After receiving the maintenance task, the maintenance personnel of each department shall respond to the maintenance task in a timely manner and carry out on-site maintenance;

[0061] S8: On-site maintenance personnel update the maintenance tasks of the Pingyuntai data center based on the actual maintenance situation.

[0062] The camera data includes a plurality of surround view cameras and a plurality of perimeter view cameras providing a large number of redundant video images to ensure coverage;

[0063] In step S3: a target detection model is constructed by a deep learning algorithm, and target detection is performed on the camera image data. The deep learning model is integrated through training of a large amount of annotated data to complete the detection of problems such as manhole cover displacement, foreign objects on the road surface, cracks in the road surface, traffic markings and damaged signs; the deep learning algorithm has good scalability and portability. Specifically, for labeling and anomaly annotation of daily collected data, it is necessary to annotate abnormal targets and target types in the picture, and use labeled image data to train a vision-based target detection deep learning model; a target detection unit is constructed for target detection during the driving process of the autonomous driving vehicle, and the target object (the category such as street lights, traffic signs, traffic markings, etc.) and anomaly boxes are marked (the anomaly box is used to indicate the detected abnormal results, such as the damaged position in the traffic sign, the displacement position of the manhole cover), and a detection unit is constructed by training a deep learning model for detecting road damage, road cracks, and foreign objects on the road surface.

[0064] In step S4: when damage or foreign objects are detected, the vehicle side organizes the vehicle information and uploads it to the cloud platform. The vehicle information includes the current positioning information of the autonomous driving vehicle, the target detection results of the deep learning model, and the image information collected by the camera module.

[0065] In step S6, after receiving the uploaded information, the cloud platform classifies and organizes the data according to different target detection results, determines the risk level, and automatically constructs maintenance tasks and assigns them to different maintenance personnel.

[0066] In an embodiment of the present invention, a method for intelligent inspection of urban roads based on an autonomous driving vehicle comprises the following steps:

[0067] S1: Obtain the location information of the autonomous driving vehicle through the positioning module installed on the vehicle;

[0068] S2: The camera module mounted on the vehicle side is used to sense the surrounding environment of the vehicle to construct image information;

[0069] S3: Using a deep learning algorithm, based on the constructed image information, perform target detection and determine whether there is a road abnormality. If yes, proceed to step S4, otherwise proceed to step S1;

[0070] S4: The vehicle side organizes the vehicle information and uploads the vehicle information to the cloud platform data center;

[0071] S5: After receiving the corresponding information, the cloud platform data center summarizes and organizes the data, stores the data in the database, builds maintenance tasks and tracks them.

[0072] Furthermore, in step S1, the positioning module includes GNSS and IMU sensor devices.

[0073] Furthermore, step S1 includes: calculating the vehicle's three-axis attitude angle and acceleration through IMU; calculating the vehicle's absolute position information through GNSS; and calculating the vehicle's high-frequency precise position information by fusing the high-frequency three-axis attitude angle and acceleration information provided by IMU and the absolute position information provided by GNSS.

[0074] Furthermore, in step S3, the daily collected data is labeled and anomalies are marked. The abnormal targets and target types in the pictures need to be marked, and the labeled image data is used to train a vision-based target detection deep learning model.

[0075] First, a 2D target detection unit is constructed for target detection during the driving process of the autonomous driving vehicle to mark the target objects and abnormal boxes.

[0076] Secondly, a multi-view BEV 3D target detection network is constructed through a zero-short method. The network inputs the multi-view image and the 2D abnormal target box or point coordinates (single point or multiple points) output by the above-mentioned 2D detection unit, and then directly outputs the position, contour, etc. of the road abnormality target in the vehicle body coordinate system under the BEV perspective through feature extraction, depth estimation, view conversion (perspective perspective to BEV perspective) and task decoder.

[0077] Furthermore, in step S3, the abnormal road conditions include: displacement of manhole covers, foreign objects on the road surface, cracks on the road surface, and damage to traffic markings and signs.

[0078] Furthermore, in step S4, the self-vehicle information includes the current position information of the autonomous driving vehicle, the target detection result of the target detection model, and the image information; and the vehicle end uploads the self-vehicle information to the cloud platform data center through the communication module.

[0079] Furthermore, in step S5, after receiving the uploaded information, the cloud platform data center classifies and organizes the data according to the target detection results, determines the risk level, and automatically constructs maintenance tasks and assigns them to different maintenance personnel.

[0080] Furthermore, in step S4, the cloud platform filters and analyzes the confidence of road anomalies at specific locations based on multiple reports from multiple vehicles and geographical locations. When the confidence is high, the risk level is determined based on the category and geographical location. When the confidence is low, the information is transferred to manual confirmation.

[0081] Furthermore, the method also includes the following steps: S6: After the maintenance personnel perform on-site maintenance, the maintenance tasks of the cloud platform data center are updated according to the actual maintenance situation.

[0082] In another embodiment of the present invention, an urban road intelligent inspection system based on an autonomous driving vehicle includes: an autonomous driving vehicle and a cloud platform data center, wherein the autonomous driving vehicle is equipped with a positioning module, a camera module and a communication module, wherein the system performs urban road intelligent inspection according to the method described above.

[0083] The implementation of the technical solution of the present invention will produce a number of beneficial effects, optimize the maintenance and management of urban infrastructure, and improve the quality of life of residents and the efficiency of urban operations. The details are as follows:

[0084] 1. High-frequency inspection and comprehensive monitoring: Autonomous driving vehicles can conduct all-weather and all-round inspections of urban roads to ensure comprehensive coverage of the infrastructure being inspected. Equipped with advanced sensor technology, they can monitor abnormal information in real time, effectively improving the timeliness and accuracy of problem discovery.

[0085] 2. Efficient response and rapid maintenance: Autonomous vehicles can achieve efficient alarms and information feedback through intelligent systems, allowing relevant maintenance personnel to respond quickly, significantly shortening the time from problem discovery to repair.

[0086] 3. Easy scalability: When there are new inspection requirements, such as adding a new detection of whether the sewer outlet is blocked, it can be supported by upgrading the new deep learning algorithm.

[0087] 4. Reduce maintenance costs: The intelligent inspection system of autonomous vehicles reduces maintenance costs by reducing dependence on manpower. The intelligent system can accurately identify the type and severity of problems, provide maintenance personnel with targeted work plans, and improve work efficiency.

[0088] 5. Data analysis and predictive maintenance: Through the accumulation and analysis of inspection data, the system can form a health status report of the infrastructure, providing a scientific basis for future maintenance work. Based on historical data, the system can also predict possible problems, making maintenance work more planned and predictive.

[0089] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable ordinary technicians in the field to understand the content of the present invention and implement it. They cannot be used to limit the protection scope of the present invention. All equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent urban road inspection method based on an autonomous driving vehicle, characterized in that: The steps include: S1: Obtain the location information of the autonomous driving vehicle through the positioning module installed on the vehicle; S2: The camera module mounted on the vehicle side is used to sense the surrounding environment of the vehicle to construct image information; S3: Using a deep learning algorithm, based on the constructed image information, perform target detection and determine whether there is a road abnormality. If yes, proceed to step S4, otherwise proceed to step S1; S4: The vehicle side organizes the vehicle information and uploads the vehicle information to the cloud platform data center; S5: After receiving the corresponding information, the cloud platform data center summarizes and organizes the data, stores the data in the database, builds maintenance tasks and tracks them.

2. The method according to claim 1, characterized in that In step S1, the positioning module includes GNSS and IMU sensor devices.

3. The method according to claim 2, characterized in that Step S1 includes: Calculate the vehicle's three-axis attitude angle and acceleration through IMU; Calculate the absolute position information of the vehicle through GNSS; By integrating the high-frequency three-axis attitude angle and acceleration information provided by IMU and the absolute position information provided by GNSS, the vehicle's high-frequency and precise position information is calculated.

4. The method according to claim 3, characterized in that In step S3, For daily collected data, labeling and anomaly annotation are required. The abnormal targets and target types in the images need to be annotated. The labeled image data is used to train a vision-based target detection deep learning model. The target detection unit is constructed for target detection during the driving process of the autonomous driving vehicle to mark the target objects and anomaly boxes. Construct a multi-view BEV network and output the depth of the abnormal box of the image in the vehicle body coordinate system.

5. The method according to claim 4, characterized in that In step S3, Abnormal road conditions include: displacement of manhole covers, foreign objects on the road surface, cracks in the road surface, and damaged traffic markings and signs.

6. The method according to claim 5, characterized in that In step S4, The self-vehicle information includes the current position information of the autonomous driving vehicle, the target detection result of the target detection model, and the image information; and The vehicle side uploads the vehicle information to the cloud platform data center through the communication module.

7. The method according to claim 6, characterized in that In step S5, after receiving the uploaded information, the cloud platform data center classifies and organizes the data according to the target detection results, determines the risk level, and automatically constructs maintenance tasks and assigns them to different maintenance personnel.

8. The method according to claim 7, characterized in that In step S4, the cloud platform filters and analyzes the confidence of road anomalies at specific locations based on multiple reports from multiple vehicles and geographic locations. When the confidence is high, the risk level is determined based on the category and geographic location. When the confidence is low, the information is transferred to manual confirmation.

9. The method according to claim 8, characterized in that The method further comprises the steps of: S6: After the maintenance personnel perform on-site maintenance, the maintenance tasks of the cloud platform data center are updated according to the actual maintenance situation.

10. An intelligent urban road inspection system based on an autonomous driving vehicle, characterized in that: It includes an autonomous driving vehicle and a cloud platform data center, wherein the autonomous driving vehicle is equipped with a positioning module, a camera module and a communication module, wherein the system performs intelligent inspection of urban roads according to the method described in any one of claims 1-9.

Images