Road defect inspection methods, devices, systems, electronic equipment and storage media

Abstract

This invention provides a method for road defect inspection, which involves acquiring at least one of the inspection environment and image quality of an inspection vehicle within the inspection range; determining a road defect detection task corresponding to the inspection range based on the inspection environment and image quality, wherein the road defect detection task includes a detection location; identifying a target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task; distributing the road defect detection task to the corresponding target vehicle and receiving the task result from the target vehicle; and determining the road defect inspection result of the inspection vehicle based on the task result. By using target vehicles to assist in road inspection, the accuracy of road defect detection during the inspection process is improved. Simultaneously, it eliminates the need for the inspection vehicle to re-inspect distant lanes and equipment facilities, avoiding the occupation of additional traffic resources.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a method, apparatus, electronic device, and storage medium for road defect inspection. Background Technology

[0002] With the application of artificial intelligence in the transportation sector, visual technology can now replace manual road inspection. A road inspection vehicle is a vehicle equipped with image sensors and image processors. It uses image sensors to collect road images or point cloud data for image recognition processing, thereby identifying road defects. Road inspections require the detection of various types of road defects, including pavement defects, roadside defects, and equipment / facilities damage. While inspection vehicles can improve efficiency, their inspection range is limited due to the widening of roads. They cannot detect distant lanes or equipment / facilities with low accuracy. Re-inspecting distant lanes and equipment / facilities would consume excessive traffic resources. Summary of the Invention

[0003] This invention provides a road defect inspection method to address the limitations of existing inspection vehicles, which suffer from limited inspection range and low accuracy in detecting distant lanes, damaged equipment, and other defects. During the inspection process, a corresponding road defect detection task is generated and distributed to a target vehicle based on at least one of the following: the inspection environment and the quality of the captured images. This target vehicle assists in the road inspection, improving the accuracy of road defect detection. Furthermore, it eliminates the need for the inspection vehicle to re-inspect distant lanes and equipment, thus avoiding the use of additional traffic resources.

[0004] In a first aspect, embodiments of the present invention provide a method for inspecting road defects, the method comprising:

[0005] Obtain at least one of the following: the inspection environment and the quality of the images captured by the inspection vehicle within its inspection range;

[0006] Based on at least one of the inspection environment and the quality of the captured image, a road defect detection task corresponding to the inspection range is determined, wherein the road defect detection task includes the detection location;

[0007] Based on the detection location of the road defect detection task, the target vehicle corresponding to the road defect detection task is determined within the inspection range;

[0008] The road defect detection task is distributed to the corresponding target vehicle, and the task results from the target vehicle are received.

[0009] The road defect inspection results of the inspection vehicle are determined based on the task results.

[0010] Optionally, obtaining the inspection environment of the inspection vehicle within the inspection range includes:

[0011] The inspection vehicle acquires images of the non-inspection lanes and roadside areas within the inspection range.

[0012] The lane environment within the inspection range is determined based on the image of the non-inspection lane;

[0013] The roadside environment within the inspection area is determined based on the roadside image;

[0014] The inspection environment within the inspection range is determined based on the lane environment and the roadside environment.

[0015] Optionally, obtaining the image quality captured by the inspection vehicle within the inspection range includes:

[0016] Obtain the image sequence of the inspection vehicle at various times within the inspection range;

[0017] Determine the captured image quality corresponding to each frame in the image sequence to obtain the captured image quality sequence of the image sequence;

[0018] The image quality within the inspection range is determined based on the captured image quality sequence.

[0019] Optionally, determining the road defect detection task corresponding to the inspection range based on at least one of the inspection environment and the quality of the captured image includes:

[0020] Determine whether the inspection environment meets the conditions for the first disease detection;

[0021] If the inspection environment does not meet the first disease detection conditions, then the first task conditions are determined based on the inspection environment, and the first task conditions include the first detection location.

[0022] Determine whether the quality of the captured image meets the second disease detection criteria;

[0023] If the quality of the captured image does not meet the second disease detection conditions, then the second task conditions are determined based on the quality of the captured image, and the second task conditions include the second detection position;

[0024] Based on at least one of the first task conditions and the second task conditions, determine the road defect detection task corresponding to the inspection range.

[0025] Optionally, determining the target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task includes:

[0026] Based on the detection locations of the road defect detection tasks, a distribution map of the road defect detection tasks within the inspection range is determined;

[0027] Obtain the vehicle distribution map within the inspection range at the current moment;

[0028] Based on the road defect detection task distribution map and the vehicle distribution map, the target vehicles corresponding to the road defect detection tasks are determined.

[0029] Optionally, the vehicle distribution map includes historical task scores for each vehicle. The step of determining the target vehicle corresponding to the road defect detection task based on the road defect detection task distribution map and the vehicle distribution map includes:

[0030] Based on the road defect detection task distribution map and the vehicle distribution map, candidate vehicles corresponding to the road defect detection tasks are determined;

[0031] Based on the historical task scores of the candidate vehicles, candidate vehicles corresponding to the road defect detection task are determined.

[0032] Optionally, after determining the road defect inspection results of the inspection vehicle based on the task results, the method further includes:

[0033] Determine the contribution of the task results to the road defect inspection results;

[0034] The task score value of the target vehicle is determined based on the contribution level;

[0035] The task score is returned to the target vehicle.

[0036] Secondly, embodiments of the present invention also provide a road defect inspection device, the road defect inspection device comprising:

[0037] The acquisition module is used to acquire at least one of the following: the inspection environment and the quality of the captured images within the inspection range of the inspection vehicle;

[0038] The first determining module is used to determine a road defect detection task corresponding to the inspection range based on at least one of the inspection environment and the quality of the captured image, wherein the road defect detection task includes the detection location;

[0039] The second determining module is used to determine the target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task.

[0040] The first processing module is used to distribute the road defect detection task to the corresponding target vehicle and receive the task results from the target vehicle.

[0041] The third determining module is used to determine the road defect inspection results of the inspection vehicle based on the task results.

[0042] Thirdly, embodiments of the present invention also provide a road defect inspection system, the road defect inspection system comprising: an inspection vehicle and a road inspection platform, wherein the inspection vehicle communicates with the road inspection platform via a communication protocol;

[0043] The road inspection platform is used to acquire at least one of the inspection environment and image quality of the inspection vehicle within the inspection range; based on at least one of the inspection environment and image quality, determine a road defect detection task corresponding to the inspection range, the road defect detection task including a detection location; determine the target vehicle corresponding to the road defect detection task within the inspection range according to the detection location of the road defect detection task; distribute the road defect detection task to the corresponding target vehicle and receive the task result from the target vehicle; and determine the road defect inspection result of the inspection vehicle based on the task result.

[0044] The inspection vehicle is used to take pictures during the inspection process and upload the pictures to the road inspection platform.

[0045] Fourthly, embodiments of the present invention provide an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the road defect inspection method provided in embodiments of the present invention.

[0046] Fifthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the road defect inspection method provided in the embodiments of the present invention.

[0047] In this embodiment of the invention, at least one of the inspection environment and the image quality of the inspection vehicle within the inspection range is acquired; based on at least one of the inspection environment and the image quality, a road defect detection task corresponding to the inspection range is determined, the road defect detection task including a detection location; according to the detection location of the road defect detection task, a target vehicle corresponding to the road defect detection task is determined within the inspection range; the road defect detection task is distributed to the corresponding target vehicle, and the task result of the target vehicle is received; based on the task result, the road defect inspection result of the inspection vehicle is determined. During the process of the inspection vehicle performing the inspection task, a corresponding road defect detection task is generated and distributed to the corresponding target vehicle based on at least one of the inspection environment and the image quality. The target vehicle assists in road inspection, improving the accuracy of road defect detection during the inspection process. Simultaneously, the inspection vehicle does not need to re-inspect distant lanes or equipment facilities, avoiding the occupation of additional traffic resources. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart of a road defect inspection method provided in an embodiment of the present invention;

[0050] Figure 2 This is a schematic diagram of the structure of a road defect inspection device provided in an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] This invention provides a road defect inspection method integrated into a road defect inspection system. The system includes at least an inspection vehicle and a road inspection platform. The inspection vehicle and the platform communicate via a vehicle-to-everything (V2X) protocol. The platform stores vehicle information for connected vehicles, including vehicle identification, driving trajectory, vehicle parameters, equipment parameters, historical task scores, and task points. The platform receives real-time information from the inspection vehicle during inspections, generates corresponding road defect detection tasks based on this information, and distributes these tasks to connected third-party vehicles. These vehicles then execute the road defect detection tasks and return the results to the platform, thus assisting the inspection vehicle in its inspection work.

[0054] The aforementioned road inspection platform can be a vehicle-to-everything (V2X) platform, and the third-party connected vehicles can communicate with the road inspection platform through the V2X protocol.

[0055] The inspection vehicle is equipped with an image capturing device that can photograph the inspected road to obtain corresponding images. It should be noted that during inspection tasks, the image capturing device can capture images at a certain frame rate, resulting in continuous images. When performing an inspection task, the inspection vehicle transmits the captured images to the road inspection platform via a communication protocol. The road inspection platform extracts at least one of the following from the received images: the inspection environment within the inspection area and the image quality. Based on at least one of the inspection environment and image quality, a road defect detection task corresponding to the inspection area is determined. The road defect detection task includes the detection location. Based on the detection location of the road defect detection task, the target vehicle corresponding to the road defect detection task is determined within the inspection area. The road defect detection task is distributed to the corresponding target vehicle, and the task results from the target vehicle are received. Based on the task results, the road defect inspection result of the inspection vehicle is determined.

[0056] The aforementioned road defects can include pavement defects and roadside defects. The aforementioned pavement defects can include pavement cracks, potholes, wavy pavement, settlement and other defects that affect pavement life and traffic safety. The aforementioned roadside defects can include equipment loss, equipment damage, facility damage, signage obstruction and signage damage and other defects that affect traffic safety.

[0057] The aforementioned inspection range can be a preset range of the current location of the inspection vehicle, such as 100 meters before and after the current location of the inspection vehicle. Alternatively, the inspection range can be the inspection section where the inspection vehicle is located after the inspection road has been divided into sections.

[0058] When the inspection environment meets the first condition for disease detection, it indicates that the inspection environment within the current inspection area is good, and assistance from third-party networked vehicles is not required. In this case, the road inspection platform can perform disease detection on the images uploaded by the inspection vehicle to obtain the disease detection results for the current inspection area. When the image quality meets the second condition for disease detection, it indicates that the image quality uploaded by the inspection vehicle is high and will not affect the disease detection results.

[0059] When the aforementioned inspection environment does not meet the conditions for detecting the first type of road defect, it indicates that the inspection environment within the current inspection area is poor, and the images uploaded by the inspection vehicle show obstructions to the road surface or roadside, making it impossible to detect road defects in these obstructed areas. This reduces the accuracy of the road defect detection results during the inspection process. In this case, the road inspection platform can determine the road defect detection task corresponding to the inspection area based on the inspection environment, identify the corresponding target vehicle based on the road defect detection task, and send the road defect detection task to the target vehicle. After receiving the road defect detection task, the target vehicle drives and takes pictures according to the task requirements, obtaining the corresponding images as the task result, and returns the task result to the road inspection platform.

[0060] For example, if within the inspection area, a vehicle is driving parallel to the inspection vehicle in a non-inspection lane next to the inspection lane, obstructing the road surface of the non-inspection lane and preventing the inspection vehicle from capturing images of the road surface, the road inspection platform can generate a road defect detection task for the corresponding non-inspection lane. Within the inspection area, a third-party vehicle in the non-inspection lane is identified as the target vehicle, and the road defect detection task for the non-inspection lane is sent to the target vehicle. After receiving the road defect detection task for the non-inspection lane, the target vehicle drives and takes images in the non-inspection lane, obtaining images of the non-inspection lane. These images are then sent to the road inspection platform as the task result. The road inspection platform performs defect detection on the images of the non-inspection lane in the task result to obtain the road defect detection result for the non-inspection lane.

[0061] For example, if the inspection environment within the inspection area is the inspection lane, not the side lane, and a vehicle is driving parallel to the inspection vehicle in the side lane, obstructing the road surface and roadside facilities, preventing the inspection vehicle from capturing images of the road surface and roadside facilities in the side lane, then the road inspection platform can generate a road defect detection task for the corresponding side lane. Within the inspection area, a third-party vehicle in the side lane is identified as the target vehicle, and the road defect detection task is sent to that target vehicle. After receiving the task, the target vehicle drives and takes images in the side lane, obtaining images of the side lane and the roadside. These images are then sent to the road inspection platform as the task result. The road inspection platform performs defect detection on the images of the side lane and the roadside in the task result, obtaining the road defect detection results for the side lane and the roadside.

[0062] When the quality of the captured images does not meet the second condition for defect detection, it indicates that the quality of the images uploaded by the inspection vehicle is low, which may affect the accuracy of the defect detection results. In this case, the road inspection platform can determine the road defect detection task corresponding to the inspection range based on the image quality, and identify the corresponding target vehicle based on the road defect detection task. The road defect detection task is then sent to the target vehicle. After receiving the road defect detection task, the target vehicle drives and takes pictures according to the task requirements, obtains the corresponding captured images as the task result, and returns the task result to the road inspection platform.

[0063] For example, if the inspection lane within the inspection range is the first lane, and the image quality of the third lane captured by the inspection vehicle does not meet the second defect detection conditions, the road inspection platform can generate a corresponding road defect detection task for the third lane. Within the inspection range, a third-party vehicle connected to the network in the third lane is identified as the target vehicle, and the road defect detection task for the third lane is sent to the target vehicle. After receiving the road defect detection task for the third lane, the target vehicle drives and takes pictures in the third lane to obtain the captured image of the third lane. The captured image of the third lane is sent to the road inspection platform as the task result. The road inspection platform performs defect detection on the captured image of the third lane in the task result to obtain the road defect detection result for the third lane.

[0064] When the inspection environment does not meet the first condition for detecting road defects and the image quality does not meet the second condition for detecting road defects, the road defect detection task corresponding to the inspection range can be determined based on the inspection environment and the image quality. The corresponding target vehicle can be identified based on the road defect detection task, and the road defect detection task can be sent to the target vehicle. After receiving the road defect detection task, the target vehicle drives and takes pictures according to the task requirements of the road defect detection task, obtains the corresponding image as the task result, and returns the task result to the road inspection platform.

[0065] In one possible embodiment, after obtaining the road defect detection results based on the task outcomes, target vehicles can be rewarded with points. Each time a vehicle completes a road defect detection task, it receives a task point value corresponding to that task. These task points can be applied in a corresponding points system. By rewarding vehicles that complete road defect detection tasks with points, the incentive for these vehicles to assist in road defect inspections can be increased.

[0066] In one possible embodiment, when a road defect detection task is sent to a target vehicle, the vehicle owner can refuse or agree to accept the task based on the actual driving situation. For example, if the target vehicle has a lane change plan within the inspection range, it can refuse to accept the corresponding road defect detection task.

[0067] In this embodiment of the invention, during the inspection process of the inspection vehicle performing the inspection task, a corresponding road defect detection task is generated based on at least one of the inspection environment and the quality of the captured image and distributed to the corresponding target vehicle. The target vehicle assists in the road inspection, improving the accuracy of road defect detection during the inspection process. At the same time, the inspection vehicle does not need to re-inspect distant lanes, equipment and facilities, thus avoiding the occupation of additional traffic resources.

[0068] like Figure 1 As shown, Figure 1 This is a flowchart of a road defect inspection method provided by an embodiment of the present invention. The road defect inspection method includes the following steps:

[0069] 101. Obtain at least one of the following: the inspection environment and the quality of the images captured by the inspection vehicle within its inspection range.

[0070] In this embodiment of the invention, the road defect inspection method described above can be mounted on a road inspection platform. This platform can be an electronic device with image processing and analysis capabilities, such as a server or server cluster. The inspection vehicle is equipped with an image capturing device, which can capture images of the inspected road to obtain corresponding images. It should be noted that when the inspection vehicle performs its inspection task on the road, the image capturing device captures images at a certain frame rate to obtain continuous images.

[0071] The aforementioned inspection range can be a preset range of the current location of the inspection vehicle, such as 100 meters before and after the current location of the inspection vehicle. Alternatively, the inspection range can be the inspection section where the inspection vehicle is located after the inspection road has been divided into sections.

[0072] The aforementioned inspection environment includes the lane environment and the roadside environment. The lane environment, also known as the road surface environment, is used to characterize whether vehicles obstruct the lane surface. The roadside environment is used to characterize whether vehicles obstruct the bypass space.

[0073] The aforementioned inspection vehicle is also equipped with environmental sensors. These sensors collect environmental information about the vehicle's surroundings in real time and transmit it to the road inspection platform. The road inspection platform then determines the inspection environment based on the received environmental information.

[0074] The image capturing equipment on the patrol vehicle captures images of the area around the vehicle in real time and transmits these images to the road patrol platform. The road patrol platform determines the patrol environment of the patrol vehicle based on the received images. Specifically, vehicle detection algorithms can be used to detect vehicles in the received images, resulting in vehicle detection results. These results include vehicle detection boxes (x1, y1, w1, h1, u1), where (x1, y1) represents the center position of the vehicle detection box, w1 represents the width of the vehicle detection box, h1 represents the width of the vehicle detection box, and u1 represents the confidence level of the vehicle detection box. The center position of the vehicle detection box can be used to determine the vehicle's location, the lane the vehicle is in can be determined based on the vehicle's location, and the patrol environment of the patrol vehicle can be determined based on the lanes of the vehicles in the image.

[0075] The aforementioned image quality is used to characterize the quality of images captured by the image capturing equipment on the inspection vehicle. When performing inspection tasks on the inspection road, the image capturing equipment captures images of the inspection road in real time, obtaining images of the inspection road, and transmits these images to the road inspection platform in real time. The road inspection platform calculates the image quality score of the inspection road images using an image quality scoring algorithm, thus obtaining the corresponding image quality for the inspection road. The aforementioned image quality scoring algorithm can be based on algorithms such as PSNR (Peak Signal-to-Noise Ratio), SSIM (Structural Similarity Index), and MSE (Mean Squared Error).

[0076] 102. Based on at least one of the inspection environment and the quality of the captured images, determine the road defect detection task corresponding to the inspection range.

[0077] In this embodiment of the invention, the road defect detection task can be understood as a task to assist in road defect inspection. The road defect detection task may include a detection location, which may be a lane that needs to be inspected. For example, the detection location may be a non-inspection lane within the inspection range.

[0078] It should be noted that, in order to ensure the accuracy of road defect inspection results, the road inspection platform will only perform defect detection on the images captured by the inspection vehicle when the inspection environment meets the requirements, or when the image quality meets the requirements, or when both the inspection environment and the image quality meet the requirements. When the inspection environment does not meet the requirements, or when the image quality does not meet the requirements, the road inspection platform will use third-party connected vehicles to collect images that meet the requirements for defect detection, thereby ensuring the accuracy of the road defect inspection results.

[0079] Specifically, when the inspection environment does not meet the requirements, it indicates that vehicles in non-inspection lanes are affecting the image capturing equipment's ability to capture images of the road surface or roadside, making it impossible to capture complete road surface or roadside images. Therefore, it is impossible to determine whether road defects exist in the obscured areas based on incomplete images. Based on this inspection environment, the detection location is determined. This detection location can be a lane within the inspection range, and capturing images of that lane can be used for road defect detection. When the detection location is a lane near the edge of the road, capturing images of both the lane and the roadside within the inspection range can be used for road defect detection.

[0080] When the quality of the captured images does not meet the requirements, the accuracy of road defect detection will decrease. The detection location can be determined based on the lanes in the corresponding images. These detection locations are the lanes within the inspection range, and capturing images of these lanes within the inspection range can be used as the task for road defect detection.

[0081] When the inspection environment does not meet the requirements and the image quality does not meet the requirements, at least one detection location can be determined by combining the position of the lane that is blocked in the inspection environment and the lane in the image. Each lane within the inspection range corresponding to each detection location is treated as a road defect detection task.

[0082] 103. Based on the detection location of the road defect detection task, determine the target vehicle corresponding to the road defect detection task within the inspection range.

[0083] In this embodiment of the invention, the detection location corresponds to a lane within the inspection range. Therefore, each road defect detection task can correspond to a lane within the inspection range, which is equivalent to a third-party vehicle driving once in a lane within the inspection range and taking pictures. The target vehicle is a third-party vehicle driving in the lane corresponding to the road defect detection task.

[0084] Specifically, after determining the road defect detection task, the road inspection platform can identify the target vehicles for each road defect detection task based on the corresponding detection location and the locations of various third-party vehicles within the inspection range. For example, third-party vehicles whose location matches the detection location can be identified as target vehicles for the corresponding road defect detection task. More specifically, third-party vehicles whose lane falls within the inspection range can be identified as target vehicles for the corresponding road defect detection task.

[0085] 104. Distribute road defect detection tasks to the corresponding target vehicles and receive the task results from the target vehicles.

[0086] In this embodiment of the invention, after identifying the target vehicles corresponding to each road defect detection task, the road inspection platform can distribute each road defect detection task to the corresponding target vehicle. Upon receiving the corresponding road defect detection task, the vehicle owner can choose to reject or accept it based on their driving situation and return the corresponding selection result to the road inspection platform. For rejected road defect detection tasks, the road inspection platform can re-identify the target vehicles within the inspection range and send the rejected road defect detection tasks to the newly identified target vehicles until the road defect detection task distribution is complete.

[0087] After the target vehicle selects to accept the corresponding road defect detection task, it can drive and take pictures according to the task requirements, obtain the corresponding pictures as the task results, and return the task results to the road inspection platform.

[0088] 105. Determine the road defect inspection results of the patrol vehicle based on the task results.

[0089] In this embodiment of the invention, the above-mentioned task result may be an image captured by the target vehicle when performing a road defect detection task. The target vehicle may return the captured image to the road inspection platform in real time during the road defect detection task, or it may return the captured image to the road inspection platform after the road defect detection task is completed.

[0090] After receiving the task results, the road inspection platform can use the trained defect detection model to detect defects in the captured images in the task results, obtain the defect detection results within the inspection range, and merge the defect detection results of each inspection range to obtain the road defect inspection results of the inspection vehicle.

[0091] The above-mentioned trained disease detection model is a disease detection model deployed in the road inspection platform. It can be trained using a dataset to obtain a trained disease detection model.

[0092] The dataset includes sample images and road defect labels corresponding to the sample images. The road defect labels corresponding to the sample images can be obtained by relevant professionals. One sample image can correspond to a set of road defect labels. A set of road defect labels includes labeling information such as defect type, defect location, and defect severity.

[0093] During training, sample images are input into the disease detection model to be trained for processing, and the disease detection results corresponding to the sample images are output. The error loss between the disease detection results corresponding to the sample images and the road disease labels corresponding to the sample images is calculated. With minimizing the error loss as the optimization objective, the parameters of the disease detection model to be trained are adjusted through the error backpropagation algorithm. The above error calculation and parameter adjustment process is iterated until the disease detection model to be trained converges at the point of minimum error loss or the number of iterations reaches the preset number, and the training is completed, resulting in a trained disease recognition model.

[0094] The trained disease detection model outputs a disease detection box (x2, y2, w2, h2, r2, u2, v2), where (x2, y2) represents the center position of the disease detection box, w2 represents the width of the disease detection box, h2 represents the height of the disease detection box, r2 represents the confidence level of the disease detection box, u represents the type of disease in the disease detection box, and v represents the degree of disease in the disease detection box.

[0095] The above-mentioned types of defects can be determined by the u in the defect detection box. These types of defects may include road surface cracking, potholes, wavy road surface, settlement, missing equipment, damaged equipment, damaged facilities, obscured signs, and damaged signs.

[0096] The aforementioned defect location can be the position of the defect in the captured image, which can be represented by the center position of the defect detection box. In one possible embodiment, each captured image has a corresponding capture time and capture position. After determining the capture position of the captured image, the defect location in the captured image can be converted into the defect location in the real world using the transformation relationship between the camera coordinate system of the image capturing device and the real-world coordinate system. In one possible embodiment, the position of the lane or road facility in the captured image can be identified simultaneously. Since the lane or road facility is known in the real world, the relative position between the defect location and the lane or road facility in the captured image can be calculated, and the corresponding defect location can be mapped to the real world based on this relative position. Specifically, since the lane or road facility in the real world is known, when the position of the lane or road facility in the captured image is detected, the mapping relationship between the position of the lane or road facility in the real world and the position of the lane or road facility in the captured image can be calculated. Through this mapping relationship, the defect location in the captured image is mapped to the real world to obtain the defect location in the real world.

[0097] The severity of the disease can be determined using the "v" indicator in the disease detection box. The severity can be expressed as a score or a grade. A higher score indicates a more severe disease, or vice versa.

[0098] In this embodiment of the invention, at least one of the inspection environment and the quality of the captured images within the inspection range of the inspection vehicle is acquired; based on at least one of the inspection environment and the captured image quality, a road defect detection task corresponding to the inspection range is determined, the road defect detection task including the detection location; according to the detection location of the road defect detection task, a target vehicle corresponding to the road defect detection task is determined within the inspection range; the road defect detection task is distributed to the corresponding target vehicle, and the task result of the target vehicle is received; the road defect inspection result of the inspection vehicle is determined based on the task result. During the process of the inspection vehicle performing the inspection task, a corresponding road defect detection task is generated based on at least one of the inspection environment and the captured image quality and distributed to the corresponding target vehicle. The target vehicle assists in road inspection, improving the accuracy of road defect detection during the inspection process. Simultaneously, the inspection vehicle does not need to re-inspect distant lanes or equipment facilities, avoiding the occupation of additional traffic resources.

[0099] It is understood that, in the specific embodiments of this application, the data related to captured images, vehicle information, etc., when applied to specific products or technologies, requires permission or consent from relevant departments, and the collection, use, and processing of the relevant data must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0100] Optionally, in the step of obtaining the inspection environment of the inspection vehicle within the inspection range, images of the non-inspection lane and roadside within the inspection range can be obtained; the lane environment within the inspection range can be determined based on the non-inspection lane images; the roadside environment within the inspection range can be determined based on the roadside images; and the inspection environment within the inspection range can be determined based on the lane environment and the roadside environment.

[0101] In this embodiment of the invention, when the inspection vehicle performs an inspection task, the image capturing device captures images of the inspection lane, the non-inspection lane, and the roadside. The inspection lane image is obtained by the image capturing device photographing the road surface of the inspection lane, which is the lane where the inspection vehicle is located. The non-inspection lane image is obtained by the image capturing device photographing the road surface of the non-inspection lane, which is the lane where the inspection vehicle is not present. The roadside image is obtained by the image capturing device photographing the space outside the lane. Multiple image capturing devices can be installed on the inspection vehicle to capture images of the inspection lane, the non-inspection lane, and the roadside separately. It is understood that the robustness of the image capturing device photographing the inspection lane is higher than that of the image capturing device photographing the non-inspection lane and the roadside. Of course, the image capturing device can also be a wide-angle or panoramic image capturing device, capable of simultaneously capturing images of the inspection lane, the non-inspection lane, and the roadside.

[0102] Generally, when inspection vehicles travel in the inspection lane, the road surface is not obstructed by other vehicles. Therefore, the inspection environment and image quality corresponding to the inspection lane images are generally good. However, because inspection vehicles travel in the inspection lane, other vehicles may travel in the non-inspection lanes, obstructing the road surface and making the road surface in the non-inspection lane images incomplete. Furthermore, the non-inspection lanes are some distance from the inspection vehicle, which may result in lower image quality. Therefore, the inspection environment and image quality corresponding to the non-inspection lane images may not meet the requirements of the inspection task, potentially leading to false detections. Similarly, because inspection vehicles travel in the inspection lane, some roadside facilities may be obstructed by vehicles in the non-inspection lanes, making the roadside facilities in the roadside images incomplete or absent. Alternatively, the inspection lane may be far from the roadside facilities, potentially resulting in lower image quality. Therefore, the inspection environment and image quality corresponding to the roadside images may not meet the requirements of the inspection task, potentially leading to false detections. Therefore, after receiving images of the patrol lane, non-patrol lane, and roadside from the patrol vehicle, the road inspection platform can determine the patrol environment within the patrol range using only the non-patrol lane and roadside images.

[0103] After receiving images of non-inspection lanes within the inspection range from the inspection vehicle, the road inspection platform can perform vehicle detection on these images to determine the presence of vehicles in the non-inspection lanes and thus obtain the lane environment within the inspection range. It can also perform vehicle detection on the non-inspection lane images to determine the density of vehicles in the non-inspection lanes. Vehicle detection algorithms can be used to perform vehicle detection on the non-inspection lane images. These algorithms can be based on Adaboost's Haar feature classifier, deep learning techniques, Gaussian mixture models, and background subtraction, among others.

[0104] After receiving roadside images from the inspection vehicle within its inspection range, the road inspection platform can perform occlusion detection on the roadside images to determine whether the roadside space is obstructed, thus obtaining the roadside environment within the inspection range. Occlusion detection algorithms can be used to perform occlusion detection on the roadside images. These algorithms can be based on background modeling and foreground segmentation techniques, deep learning techniques, shape matching, and feature descriptor techniques, among others.

[0105] After obtaining the lane environment and roadside environment, the inspection environment within the inspection range can be determined based on whether there are vehicles driving in the non-inspection lane and whether the roadside space is obstructed.

[0106] Optionally, in the step of obtaining the image quality of the patrol vehicle within the patrol range, the image sequence corresponding to each moment of the patrol vehicle within the patrol range can be obtained; the image quality corresponding to each frame in the image sequence can be determined to obtain the image quality sequence of the image sequence; and the image quality within the patrol range can be determined based on the image quality sequence.

[0107] In this embodiment of the invention, when the inspection vehicle is performing an inspection task, the image capturing device can capture images at a certain frame rate to obtain continuous images. Furthermore, when the inspection vehicle is performing an inspection task, the image capturing device captures images at a certain frame rate within the inspection range to obtain continuous images within the inspection range. These continuous images within the inspection range are determined as the image sequence corresponding to each moment within the inspection range and uploaded to the road inspection platform.

[0108] Upon obtaining the aforementioned image sequence, the road inspection platform can calculate the image quality score for each frame using an image quality scoring algorithm. This score determines the overall image quality of the corresponding frame, and based on the image quality of each frame, a corresponding image quality sequence is determined. The image quality scoring algorithm can be based on algorithms such as PSNR (Peak Signal-to-Noise Ratio), SSIM (Structural Similarity Index), or MSE (Mean Squared Error).

[0109] After obtaining the image quality sequence corresponding to the aforementioned image sequence, the image quality within the inspection range can be determined based on the average image quality score of the image quality sequence. Alternatively, the image quality sequence can be input into a trained temporal prediction network for processing, outputting a predicted image quality score, which can then be used to determine the image quality within the inspection range. The aforementioned temporal prediction network can be based on an autoregressive moving average model (ARMA), a long short-term memory neural network (LSTM), or a Kalman filter, among others.

[0110] Optionally, in the step of determining the road defect detection task corresponding to the inspection range based on at least one of the inspection environment and the quality of the captured image, it can be determined whether the inspection environment meets the first defect detection condition; if the inspection environment does not meet the first defect detection condition, then the first task condition is determined based on the inspection environment, the first task condition including the first detection location; it can be determined whether the quality of the captured image meets the second defect detection condition; if the captured image quality does not meet the second defect detection condition, then the second task condition is determined based on the captured image quality, the second task condition including the second detection location; and the road defect detection task corresponding to the inspection range is determined based on at least one of the first task condition and the second task condition.

[0111] In this embodiment of the invention, when the inspection environment meets the first defect detection condition, it indicates that the inspection environment within the current inspection range is good, and assistance from third-party networked vehicles is not required. In this case, the road inspection platform can perform defect detection on the images uploaded by the inspection vehicle to obtain the defect detection results for the current inspection range. When the image quality meets the second defect detection condition, it indicates that the image quality uploaded by the inspection vehicle is high and will not affect the defect detection results.

[0112] Furthermore, the aforementioned inspection environment includes the lane environment and the roadside environment. The first defect detection condition can be whether there are vehicles traveling on the lane or whether the roadside space is obstructed. If there are no vehicles traveling in the lane environment, the corresponding lane environment meets the first defect detection condition; if there are vehicles traveling in the lane environment, the corresponding lane environment does not meet the first defect detection condition. Similarly, if the roadside space is not obstructed in the roadside environment, the corresponding roadside environment meets the first defect detection condition; if the roadside space is obstructed in the roadside environment, the corresponding roadside environment does not meet the first defect detection condition. The aforementioned first judgment detection condition can also be whether the density of vehicles traveling on the lane is less than a density threshold. If the density of vehicles traveling in the lane environment is greater than or equal to the density threshold, the corresponding lane environment does not meet the first defect detection condition; if the density of vehicles traveling in the lane environment is less than the density threshold, the corresponding lane environment meets the first defect detection condition. Furthermore, the aforementioned lane environment refers to the environment of a non-inspection lane. The aforementioned first defect detection condition can be whether there are vehicles traveling in the non-inspection lane, or whether the density of vehicles traveling in the non-inspection lane is less than a density threshold. When there are vehicles traveling in the non-inspection lane environment, or the density of vehicles traveling in the lane environment is greater than or equal to the density threshold, it can be said that the corresponding lane environment does not meet the first defect detection condition. When there are no vehicles traveling in the non-inspection lane environment, or the density of vehicles traveling in the lane environment is less than the density threshold, it can be said that the corresponding lane environment meets the first defect detection condition.

[0113] When the aforementioned inspection environment does not meet the conditions for detecting the first type of road defect, it indicates that the inspection environment within the current inspection area is poor, and the images uploaded by the inspection vehicle show obstructions to the road surface or roadside, making it impossible to detect road defects on the obstructed road surface or roadside, thus reducing the accuracy of the road defect detection results during the inspection process. In this case, the road inspection platform can determine the road defect detection task corresponding to the inspection area based on the inspection environment.

[0114] For example, if the inspection environment within the inspection area involves a vehicle driving parallel to the inspection vehicle in a non-inspection lane next to the inspection lane, obstructing the inspection vehicle's view of the non-inspection lane's road surface and preventing the inspection vehicle from capturing images of the lane's road condition, the road inspection platform can generate a first task condition for the corresponding non-inspection lane and determine the road defect detection task corresponding to the inspection area based on this condition. Furthermore, the first task condition includes a first detection location and a first detection target. The first detection location can be the corresponding non-inspection lane, and the first detection target can be the road surface. The corresponding road defect detection task is determined based on the first detection location and the first detection target.

[0115] For example, if the inspection environment within the inspection area is the inspection lane, not the side lane, and a vehicle is driving parallel to the inspection vehicle in the side lane, obstructing the view of the road surface and roadside facilities, preventing the inspection vehicle from capturing images of the road surface and facilities in the side lane, then the road inspection platform can generate a first task condition and determine the corresponding road defect detection task within the inspection area based on this condition. Furthermore, the first task condition includes a first detection location and a first detection target. The first detection location can be the corresponding non-inspection lane, and the first detection target can be the roadside. The corresponding road defect detection task is determined based on the first detection location and the first detection target.

[0116] If the quality of the captured images does not meet the second condition for defect detection, it indicates that the quality of the images uploaded by the inspection vehicle is low, which may affect the accuracy of the defect detection results. In this case, the road inspection platform can determine the road defect detection task corresponding to the inspection area based on the quality of the captured images.

[0117] Furthermore, the aforementioned image quality can be the average image quality score of the image sequence, or the predicted image quality of the image sequence. The aforementioned second disease detection condition can be whether the average image quality score is greater than or equal to the first scoring threshold, or whether the predicted image quality score is greater than or equal to the second scoring threshold. When the average image quality score is greater than or equal to the first scoring threshold, it indicates that the corresponding image sequence meets the second disease detection condition. Alternatively, when the predicted image quality score is greater than or equal to the second scoring threshold, it indicates that the corresponding image sequence meets the second disease detection condition. When the average image quality score is less than the first scoring threshold, it indicates that the corresponding image sequence does not meet the second disease detection condition. Alternatively, when the predicted image quality score is less than the second scoring threshold, it indicates that the corresponding image sequence does not meet the second disease detection condition.

[0118] If the image quality does not meet the second defect detection conditions, a second task condition is determined based on the image quality. The second task condition includes a second detection location and a second detection target. Specifically, if the image quality does not meet the second defect detection conditions, the road inspection platform can generate a first task condition corresponding to the non-inspection lane, and determine the road defect detection task corresponding to the inspection range based on the first task condition. Further, the aforementioned second task condition includes a second detection location and a second detection target. The second detection location can be the corresponding non-inspection lane, and the second detection target can be the road surface or roadside. The corresponding road defect detection task is determined based on the second detection location and the second detection target.

[0119] For example, if the inspection lane within the inspection range is the first lane, and the image quality of the third lane captured by the inspection vehicle does not meet the second defect detection conditions, then the road inspection platform can generate the corresponding second task conditions for the third lane. In the second task conditions, the second detection location is the third lane, and the second task target can be the road surface of the third lane or the roadside of the third lane. The second detection location and the second detection target determine the corresponding road defect detection tasks and save them one by one.

[0120] When the inspection environment does not meet the first condition for detecting road defects, and the image quality does not meet the second condition for detecting road defects, first and second task conditions can be generated based on the inspection environment and image quality. The road defect detection task corresponding to the inspection area can then be determined based on these first and second task conditions. Alternatively, the first and second task conditions can be merged, and the road defect detection task corresponding to the inspection area can be determined based on the merged first and second task conditions.

[0121] Optionally, in the step of determining the target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task, a road defect detection task distribution map within the inspection range can be determined based on the detection location of the road defect detection task; a vehicle distribution map within the inspection range at the current moment can be obtained; and the target vehicle corresponding to the road defect detection task can be determined based on the road defect detection task distribution map and the vehicle distribution map.

[0122] In this embodiment of the invention, the road defect detection task includes a detection location and a detection target. The detection location can be a lane location, and the detection target can be a road surface or roadside. A map corresponding to the inspection range can be obtained, and the detection location corresponding to the road defect detection task can be added to the map corresponding to the inspection range to obtain a road defect detection task distribution map.

[0123] It can obtain the location of vehicles within the inspection range at the current time, add the parking space locations within the inspection range at the current time to the map corresponding to the inspection range, and obtain the vehicle distribution map within the inspection range at the current time.

[0124] Specifically, task distribution lanes can be determined in the aforementioned road defect detection task distribution map. Based on these task distribution lanes, target lanes can be determined in the vehicle distribution map, and target vehicles can be identified within those target lanes. Alternatively, vehicles that have just entered the inspection area within the target lane can be identified as target vehicles, or the slowest vehicle entering the network can be identified as the target vehicle within the target vehicle's lane.

[0125] Optionally, the vehicle distribution map includes the historical task scores of each vehicle. In the step of determining the target vehicle corresponding to the road defect detection task based on the road defect detection task distribution map and the vehicle distribution map, candidate vehicles corresponding to the road defect detection task can be determined based on the road defect detection task distribution map and the vehicle distribution map; and candidate vehicles corresponding to the road defect detection task can be determined based on the historical task scores of the candidate vehicles.

[0126] In this embodiment of the invention, the aforementioned historical task score represents the vehicle's performance in completing road defect detection tasks over a past period. A higher historical task score indicates a higher degree of completion of the road defect detection task. The historical task score can be determined based on the image quality of the images captured in each task result; higher image quality in the task result corresponds to a higher task score.

[0127] It can obtain the location of vehicles within the inspection range at the current moment, add the parking space location and corresponding vehicle information within the inspection range to the map corresponding to the inspection range at the current moment, and obtain the vehicle distribution map within the inspection range at the current moment. The vehicle distribution map includes vehicle information, such as vehicle identification, vehicle driving trajectory, vehicle parameters, equipment parameters, historical task scores and task points.

[0128] The task distribution lanes are determined in the above road defect detection task distribution map. Based on the task distribution lanes, the target lanes are determined in the vehicle distribution map. Multiple vehicles entering the inspection range in the target lanes are identified as candidate vehicles. The candidate vehicle with the highest historical task score is identified as the target vehicle.

[0129] Optionally, after determining the road defect inspection results of the inspection vehicle based on the task results, the contribution of the task results to the road defect inspection results can also be determined; the task score of the target vehicle can be determined based on the contribution; and the task score can be returned to the target vehicle.

[0130] In this embodiment of the invention, after receiving the task results, the road inspection platform can use a trained defect detection model to perform defect detection on the captured images in the task results, obtain the defect detection results within the inspection range, and merge the defect detection results within each inspection range to obtain the road defect inspection results of the inspection vehicle.

[0131] The contribution of the above task results to the road defect inspection results refers to the importance of the defect detection results corresponding to the task results in the road defect inspection results. The contribution of the above task results to the road defect inspection results can be calculated by the following formula:

[0132]

[0133] Among them, the above This represents the contribution of the i-th task result to the road defect inspection result, as stated above. Indicates the first i The first task result j The confidence level corresponding to each disease detection result. Indicates the first i The first task result j The disease type corresponding to each disease detection result. Indicates the first i The first task result j The degree of disease corresponding to each disease detection result. n Indicates the number of task results. m This represents the number of disease detection results in the i-th task result.

[0134] After determining the contribution of the task results to the road defect inspection results, the task points value of the target vehicle can be determined based on the contribution; the greater the contribution, the higher the corresponding task points value. After obtaining the target vehicle's task points value, the task points value in the target vehicle's vehicle information is updated accordingly. Simultaneously, the target vehicle's task points value is sent to the target vehicle so that the vehicle owner is aware of the task points value earned from performing this road defect inspection task. By rewarding target vehicles with points, each completed road defect inspection task earns a corresponding task points value, which can be applied in the corresponding points system. Rewarding networked vehicles for completing road defect inspection tasks with points can increase their enthusiasm for assisting in road defect inspections.

[0135] It should be noted that the road defect inspection method provided in this embodiment of the invention can be applied to devices such as camera equipment, smartphones, computers, and servers that can perform road defect inspections.

[0136] like Figure 2 As shown, an embodiment of the present invention provides a road defect inspection device, which includes:

[0137] The acquisition module 201 is used to acquire at least one of the inspection environment and the quality of the captured images within the inspection range of the inspection vehicle;

[0138] The first determining module 202 is used to determine a road defect detection task corresponding to the inspection range based on at least one of the inspection environment and the quality of the captured image, wherein the road defect detection task includes a detection location;

[0139] The second determining module 203 is used to determine the target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task.

[0140] The first processing module 204 is used to distribute the road defect detection task to the corresponding target vehicle and receive the task results from the target vehicle.

[0141] The third determining module 205 is used to determine the road defect inspection results of the inspection vehicle based on the task results.

[0142] Optionally, the acquisition module 201 includes:

[0143] The first acquisition submodule is used to acquire images of the non-inspection lanes and roadside images of the inspection vehicle within the inspection range;

[0144] The first determining submodule is used to determine the lane environment within the inspection range based on the non-inspection lane image;

[0145] The second determining submodule is used to determine the roadside environment within the inspection range based on the roadside image;

[0146] The third determining submodule is used to determine the inspection environment within the inspection range based on the lane environment and the roadside environment.

[0147] Optionally, the acquisition module 201 includes:

[0148] The second acquisition submodule is used to acquire the image sequence corresponding to each moment of the inspection vehicle within the inspection range;

[0149] The fourth determining submodule is used to determine the captured image quality corresponding to each frame of the image sequence, so as to obtain the captured image quality sequence of the image sequence;

[0150] The fifth determining submodule is used to determine the image quality within the inspection range based on the captured image quality sequence.

[0151] Optionally, the first determining module 202 includes:

[0152] The first judgment submodule is used to determine whether the inspection environment meets the first disease detection conditions;

[0153] The sixth determining submodule is used to determine the first task conditions based on the inspection environment if the inspection environment does not meet the first disease detection conditions. The first task conditions include the first detection location.

[0154] The second judgment submodule is used to determine whether the quality of the captured image meets the second disease detection conditions;

[0155] The seventh determination submodule is used to determine the second task conditions based on the image quality if the quality of the captured image does not meet the second disease detection conditions. The second task conditions include the second detection position.

[0156] The eighth determination submodule is used to determine the road defect detection task corresponding to the inspection range based on at least one of the first task conditions and the second task conditions.

[0157] Optionally, the second determining module 203 includes:

[0158] The ninth determining submodule is used to determine the distribution map of road defect detection tasks within the inspection range based on the detection location of the road defect detection tasks;

[0159] The third acquisition submodule is used to acquire the vehicle distribution map within the inspection range at the current moment;

[0160] The tenth determination submodule is used to determine the target vehicle corresponding to the road defect detection task based on the road defect detection task distribution map and the vehicle distribution map.

[0161] Optionally, the tenth determining submodule includes:

[0162] The first determining unit is used to determine candidate vehicles corresponding to the road defect detection task based on the road defect detection task distribution map and the vehicle distribution map;

[0163] The second determining unit is used to determine the candidate vehicle corresponding to the road defect detection task based on the historical task scores of the candidate vehicles.

[0164] Optionally, the device further includes:

[0165] The fourth determining module is used to determine the contribution of the task results to the road defect inspection results;

[0166] The fifth determining module is used to determine the task score value of the target vehicle based on the contribution level;

[0167] The second processing module is used to return the task score value to the target vehicle.

[0168] It should be noted that the road defect inspection device provided in this embodiment of the invention can be applied to devices such as camera equipment, smartphones, computers, and servers that can perform road defect inspections.

[0169] The road defect inspection device provided in this embodiment of the invention can realize all the processes implemented by the road defect inspection method in the above-described method embodiments, and can achieve the same beneficial effects. To avoid repetition, it will not be described again here.

[0170] See Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 3 As shown, it includes: a memory 302, a processor 301, and a computer program for a road defect inspection method stored in the memory 302 and executable on the processor 301, wherein:

[0171] The processor 301 is used to call the computer program stored in the memory 302 and perform the following steps:

[0172] Obtain at least one of the following: the inspection environment and the quality of the images captured by the inspection vehicle within its inspection range;

[0173] Based on at least one of the inspection environment and the quality of the captured image, a road defect detection task corresponding to the inspection range is determined, wherein the road defect detection task includes the detection location;

[0174] Based on the detection location of the road defect detection task, the target vehicle corresponding to the road defect detection task is determined within the inspection range;

[0175] The road defect detection task is distributed to the corresponding target vehicle, and the task results from the target vehicle are received.

[0176] The road defect inspection results of the inspection vehicle are determined based on the task results.

[0177] Optionally, the process of obtaining the inspection environment of the inspection vehicle within the inspection range, executed by processor 301, includes:

[0178] The inspection vehicle acquires images of the non-inspection lanes and roadside areas within the inspection range.

[0179] The lane environment within the inspection range is determined based on the image of the non-inspection lane;

[0180] The roadside environment within the inspection area is determined based on the roadside image;

[0181] The inspection environment within the inspection range is determined based on the lane environment and the roadside environment.

[0182] Optionally, the process of acquiring the image quality of the inspection vehicle within the inspection range, executed by processor 301, includes:

[0183] Obtain the image sequence of the inspection vehicle at various times within the inspection range;

[0184] Determine the captured image quality corresponding to each frame in the image sequence to obtain the captured image quality sequence of the image sequence;

[0185] The image quality within the inspection range is determined based on the captured image quality sequence.

[0186] Optionally, the task of determining the road defect detection corresponding to the inspection range based on at least one of the inspection environment and the quality of the captured image, executed by the processor 301, includes:

[0187] Determine whether the inspection environment meets the conditions for the first disease detection;

[0188] If the inspection environment does not meet the first disease detection conditions, then the first task conditions are determined based on the inspection environment, and the first task conditions include the first detection location.

[0189] Determine whether the quality of the captured image meets the second disease detection criteria;

[0190] If the quality of the captured image does not meet the second disease detection conditions, then the second task conditions are determined based on the quality of the captured image, and the second task conditions include the second detection position;

[0191] Based on at least one of the first task conditions and the second task conditions, determine the road defect detection task corresponding to the inspection range.

[0192] Optionally, the step of processor 301 determining the target vehicle corresponding to the road defect detection task within the inspection range based on the detection location of the road defect detection task includes:

[0193] Based on the detection locations of the road defect detection tasks, a distribution map of the road defect detection tasks within the inspection range is determined;

[0194] Obtain the vehicle distribution map within the inspection range at the current moment;

[0195] Based on the road defect detection task distribution map and the vehicle distribution map, the target vehicles corresponding to the road defect detection tasks are determined.

[0196] Optionally, the vehicle distribution map includes historical task scores for each vehicle. The process executed by processor 301, based on the road defect detection task distribution map and the vehicle distribution map, to determine the target vehicle corresponding to the road defect detection task, includes:

[0197] Based on the road defect detection task distribution map and the vehicle distribution map, candidate vehicles corresponding to the road defect detection tasks are determined;

[0198] Based on the historical task scores of the candidate vehicles, candidate vehicles corresponding to the road defect detection task are determined.

[0199] Optionally, after determining the road defect inspection results of the inspection vehicle based on the task results, the method executed by the processor 301 further includes:

[0200] Determine the contribution of the task results to the road defect inspection results;

[0201] The task score value of the target vehicle is determined based on the contribution level;

[0202] The task score is returned to the target vehicle.

[0203] It should be noted that the electronic device provided in the embodiments of the present invention can be applied to devices such as smartphones, computers, and servers that can perform road defect inspection methods.

[0204] The electronic device provided in this embodiment of the invention can implement all the processes of the road defect inspection method in the above-described method embodiments, and can achieve the same beneficial effects. To avoid repetition, further details are omitted here.

[0205] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the road defect inspection method or the application-side road defect inspection method provided in this invention, and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0206] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0207] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A road disease inspection method characterized by comprising: The method comprises the following steps: acquiring at least one of a patrol environment and a shooting image quality of the patrol vehicle in a patrol range; determining a road disease detection task corresponding to the patrol range based on at least one of the patrol environment and the shooting image quality, the road disease detection task comprising a detection position; determining a target vehicle corresponding to the road disease detection task according to the detection position of the road disease detection task in the patrol range; distributing the road disease detection task to the corresponding target vehicle and receiving a task result of the target vehicle; determining a road disease patrol result of the patrol vehicle based on the task result; the determining of the target vehicle corresponding to the road disease detection task according to the detection position of the road disease detection task in the patrol range comprises: determining a road disease detection task distribution map in the patrol range based on the detection position of the road disease detection task; acquiring a vehicle distribution map in the patrol range at a current time; determining the target vehicle corresponding to the road disease detection task based on the road disease detection task distribution map and the vehicle distribution map.

2. The road defect inspection method according to claim 1, wherein the acquiring of the patrol environment of the patrol vehicle in the patrol range comprises: acquiring a non-patrol lane image and a roadside image of the patrol vehicle in the patrol range; determining a lane environment in the patrol range according to the non-patrol lane image; determining a roadside environment in the patrol range according to the roadside image; determining the patrol environment in the patrol range according to the lane environment and the roadside environment.

3. The road defect inspection method according to claim 2, wherein the acquiring of the shooting image quality of the patrol vehicle in the patrol range comprises: acquiring an image sequence corresponding to each time in the patrol range; determining a shooting image quality corresponding to each frame of image in the image sequence to obtain a shooting image quality sequence of the image sequence; determining the shooting image quality in the patrol range based on the shooting image quality sequence.

4. The road defect inspection method according to claim 3, wherein the determining of the road disease detection task corresponding to the patrol range based on at least one of the patrol environment and the shooting image quality comprises: judging whether the patrol environment meets a first disease detection condition; if the patrol environment does not meet the first disease detection condition, determining a first task condition according to the patrol environment, the first task condition comprising a first detection position; judging whether the shooting image quality meets a second disease detection condition; if the shooting image quality does not meet the second disease detection condition, determining a second task condition according to the shooting image quality, the second task condition comprising a second detection position; determining the road disease detection task corresponding to the patrol range based on at least one of the first task condition and the second task condition.

5. The road defect inspection method according to claim 1, wherein the vehicle distribution map comprises a historical task score of each vehicle, and the determining of the target vehicle corresponding to the road disease detection task based on the road disease detection task distribution map and the vehicle distribution map comprises: determining a candidate vehicle corresponding to the road disease detection task based on the road disease detection task distribution map and the vehicle distribution map; determine a candidate vehicle corresponding to the road disease detection task based on historical task scores of the candidate vehicle.

6. The road defect inspection method according to claim 1, wherein After determining the road disease inspection result of the inspection vehicle based on the task result, the method further comprises: determining a contribution degree of the task result to the road disease inspection result; determining a task score value of the target vehicle based on the contribution degree; returning the task score value to the target vehicle.

7. A road disease inspection device characterized by comprising: The road disease inspection device comprises: an acquisition module configured to acquire at least one of an inspection environment and a shooting image quality of an inspection vehicle in an inspection range; a first determination module configured to determine a road disease detection task corresponding to the inspection range based on at least one of the inspection environment and the shooting image quality, the road disease detection task comprising a detection position; a second determination module configured to determine a target vehicle corresponding to the road disease detection task based on the detection position of the road disease detection task in the inspection range; a first processing module configured to distribute the road disease detection task to the corresponding target vehicle and receive a task result of the target vehicle; a third determination module configured to determine a road disease inspection result of the inspection vehicle based on the task result; The second determination module is configured to determine a target vehicle corresponding to the road disease detection task based on the detection position of the road disease detection task in the inspection range, and specifically configured to: determine a road disease detection task distribution map in the inspection range based on the detection position of the road disease detection task; acquire a vehicle distribution map in the inspection range at a current time; determine a target vehicle corresponding to the road disease detection task based on the road disease detection task distribution map and the vehicle distribution map.

8. A road disease inspection system characterized by comprising: The road disease inspection system comprises an inspection vehicle and a road inspection platform, and the inspection vehicle is communicatively connected to the road inspection platform through a communication protocol. The road inspection platform is configured to acquire at least one of an inspection environment and a shooting image quality of an inspection vehicle in an inspection range; determine a road disease detection task corresponding to the inspection range based on at least one of the inspection environment and the shooting image quality, the road disease detection task comprising a detection position; determine a target vehicle corresponding to the road disease detection task based on the detection position of the road disease detection task in the inspection range; distribute the road disease detection task to the corresponding target vehicle and receive a task result of the target vehicle; and determine a road disease inspection result of the inspection vehicle based on the task result. The inspection vehicle is configured to shoot images during an inspection process and upload the shot images to the road inspection platform.

9. An electronic device, comprising: The road disease inspection device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the road disease inspection method according to any one of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps in the road disease inspection method according to any one of claims 1 to 6.

Images