Road asset detection management method and device based on vehicle-road cooperation and storage medium

Abstract

The application discloses a kind of road asset detection management methods based on vehicle-road cooperation, computer device and storage medium, including respectively through road side equipment and car end equipment to the perception of road asset, obtain first perception data and second perception data, first perception data and second perception data are synergistically optimized and integrated, obtain asset information table and the like steps.The application can perceive the road asset around the fixed position by using road side equipment, and can perceive the area that cannot be covered by road side equipment by using car end equipment, so as to expand the coverage of road asset, and a large amount of road asset can be managed digitally;On the other hand, by synergistically optimizing the first perception data obtained by road side equipment and the second perception data obtained by car end equipment, the accuracy of the second perception data can be improved, which is beneficial to the fine management of road asset.The application is widely used in the technical field of asset management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of asset management, and in particular to a road asset detection management method based on vehicle-road cooperation, a computer device and a storage medium. BACKGROUND

[0002] Road assets include road land, roads (including road bridges, road tunnels, road crossings, etc.), structures, traffic engineering and facilities along the line (including traffic safety facilities, management facilities, service facilities, green environmental protection facilities), and other assets that are indispensable components of normal use of roads. With the continuous growth of urban road infrastructure, it has brought great convenience to people's travel, while the management and maintenance of road assets also face greater challenges. At present, road maintenance and repair management still mainly relies on manual detection. Regularly organize detection and census within the jurisdiction. This kind of inspection method is difficult to remember, fine and real-time detection due to the large number of types and quantities of road assets, which can easily cause asset damage to be difficult to identify, difficult to find in time, and has problems such as large consumption of human resources and long detection cycle. SUMMARY

[0003] In view of the technical problems in the prior art that the road asset management technology is difficult to fine, resource consumption is large, and the detection cycle is long, the purpose of the present application is to provide a road asset detection management method based on vehicle-road cooperation, a computer device and a storage medium.

[0004] In one aspect, the present application embodiment includes a road asset detection management method based on vehicle-road cooperation, which performs a plurality of detection cycles and performs the following steps in any detection cycle:

[0005] The road assets in the installation location environment are sensed by a plurality of roadside devices to obtain a plurality of first sensing data; the first sensing data is used to represent the coordinates and attributes of the road assets sensed by the roadside device;

[0006] The road assets in the driving path environment are sensed by a vehicle-side device to obtain second sensing data; the second sensing data is used to represent the coordinates and attributes of the road assets sensed by the vehicle-side device;

[0007] The first sensing data and the second sensing data are cooperatively optimized;

[0008] The first sensing data and the second sensing data are integrated to obtain an asset information table.

[0009] Further, the road assets in the installation location environment are sensed by a plurality of roadside devices to obtain a plurality of first sensing data, which includes:

[0010] For any one of the roadside devices, a first image data is obtained by performing point cloud and image perception scanning on an environment where an installation location is located through the roadside device;

[0011] A first device coordinate data is obtained by positioning through the roadside device;

[0012] A first attribute data and a first relative coordinate data are obtained by processing the first image data using a three-dimensional target detection algorithm; the first attribute data is used to represent an attribute of a road asset in the first image data, and the first relative coordinate data is used to represent a coordinate of the road asset in the first image data relative to the roadside device;

[0013] A first absolute coordinate data is determined according to the first device coordinate data and the first relative coordinate data; the first absolute coordinate data is used to represent an absolute coordinate of the road asset in the first image data;

[0014] The first perception data is generated according to the first attribute data and the first absolute coordinate data.

[0015] Further, a second perception data is obtained by performing perception on road assets of an environment passed by a driving path through a vehicle-side device, including:

[0016] A coverage range of each of the roadside devices is determined according to each of the first device coordinate data;

[0017] A driving path of the vehicle-side device is planned; the driving path bypasses the coverage range of each of the roadside devices;

[0018] A second device coordinate data is obtained by positioning through the vehicle-side device;

[0019] A second image data is obtained by performing point cloud and image perception scanning on an environment passed by the driving path through the vehicle-side device along the driving path;

[0020] A second attribute data and a second relative coordinate data are obtained by processing the second image data using a three-dimensional target detection algorithm; the second attribute data is used to represent an attribute of a road asset in the second image data, and the second relative coordinate data is used to represent a coordinate of the road asset in the second image data relative to the vehicle-side device;

[0021] A second absolute coordinate data is determined according to the second device coordinate data and the second relative coordinate data; the second absolute coordinate data is used to represent an absolute coordinate of the road asset in the second image data;

[0022] According to the second attribute data and the second absolute coordinate data, the second perception data is generated.

[0023] Further, the collaborative optimization of the first perception data and the second perception data comprises:

[0024] Each of the roadside devices is respectively provided with precise coordinate data, which is used to represent the coordinate of the installation position of the corresponding roadside device.

[0025] Each of the roadside devices respectively determines differential correction data according to the corresponding precise coordinate data and the first device coordinate data obtained by positioning.

[0026] According to each of the differential correction data, the second perception data is differentially corrected.

[0027] Further, the differential correction of the second perception data according to each of the differential correction data comprises:

[0028] The environmental similarity between the vehicle-side device and each of the roadside devices is respectively determined.

[0029] For any one of the differential correction data, a corresponding weight is given to the differential correction data according to the environmental similarity between the corresponding roadside device and the vehicle-side device of the differential correction data.

[0030] According to each of the differential correction data, the second perception data is differentially corrected according to the corresponding weight of each of the differential correction data.

[0031] Further, the following steps are further performed in any one of the detection periods:

[0032] After the collaborative optimization of the first perception data and the second perception data, the first perception data and the second perception data are denoised before the integration of the first perception data and the second perception data to obtain the asset information table.

[0033] Further, the denoising of the first perception data and the second perception data comprises:

[0034] A convolutional neural network is established.

[0035] A first two-dimensional image is obtained.

[0036] Zero-mean noise is added to the first two-dimensional image to obtain a second two-dimensional image.

[0037] inputting the first two-dimensional image into the trained convolutional neural network for processing.

[0038] converting the first perception data and the second perception data into third two-dimensional images respectively;

[0039] inputting the third two-dimensional image into the trained convolutional neural network for processing.

[0040] Further, the road asset detection management method based on vehicle-road cooperation further comprises the following steps:

[0041] comparing the asset information tables obtained by performing the detection cycles respectively;

[0042] performing asset management according to the comparison result.

[0043] In another aspect, the embodiment of the present application further comprises a computer device comprising a memory and a processor, wherein the memory is used to store at least one program, and the processor is used to load the at least one program to execute the road asset detection management method based on vehicle-road cooperation in the embodiment.

[0044] In another aspect, the embodiment of the present application further comprises a storage medium having a processor-executable program stored therein, wherein the processor-executable program is used to execute the road asset detection management method based on vehicle-road cooperation in the embodiment when executed by a processor.

[0045] The road asset detection management method based on vehicle-road cooperation in the embodiment has the following beneficial effects: the road asset detection management method based on vehicle-road cooperation in the embodiment can perceive road assets around a fixed position by using roadside equipment, and can perceive areas that cannot be covered by roadside equipment by using vehicle-side equipment, thereby expanding the coverage of road assets and enabling data management of a large number of road assets; on the other hand, the first perception data obtained by roadside equipment and the second perception data obtained by vehicle-side equipment are cooperatively optimized, which can improve the accuracy of the second perception data and is conducive to fine management of road assets. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 The figure is a schematic diagram of the steps of the road asset detection management method based on vehicle-road cooperation in the embodiment;

[0047] Figure 2 The figure is a schematic diagram of the system to which the road asset detection management method based on vehicle-road cooperation in the embodiment can be applied;

[0048] Figure 3 The figure is a schematic diagram of the flow of the road asset detection management method based on vehicle-road cooperation in the embodiment;

[0049] Figure 4 Fig. 1 is a schematic diagram illustrating the principle of denoising first perception data and second perception data in an embodiment. DETAILED DESCRIPTION

[0050] In this embodiment, when performing the road asset detection management method based on vehicle-road cooperation, a plurality of detection cycles are executed in a loop, for example, after performing a detection cycle, the next detection cycle is started, and the steps performed in each detection cycle can be the same. Take one of the detection cycles as an example for illustration. Referring to Fig. 1, any one of the detection cycles includes the following steps: Figure 1

[0051] S1. A plurality of roadside devices perceive road assets in the environment where the installation position is located to obtain a plurality of first perception data; the first perception data is used to represent the coordinates and attributes of the road assets perceived by the roadside devices;

[0052] S2. A vehicle-side device perceives road assets in the environment where the driving path passes to obtain second perception data; the second perception data is used to represent the coordinates and attributes of the road assets perceived by the vehicle-side device;

[0053] S3. The first perception data and the second perception data are cooperatively optimized;

[0054] S4. The first perception data and the second perception data are integrated to obtain an asset information table.

[0055] Steps S1-S4 can be applied in the system shown in Fig. 1. Referring to Fig. 1, the system includes a plurality of roadside devices 1, a vehicle-side device 2, and a server 3. Figure 2 Figure 2 ​​The roadside device 100 is installed at a plurality of positions such as both sides of a road and an intersection, and the vehicle terminal device 200 is installed on a vehicle. The roadside device and the vehicle terminal device are similar in function and structure, and each is provided with a power supply module, a point cloud and image perception scanning module, a positioning module, a processing module, a storage module, and a communication module. The point cloud and image perception scanning module scans and perceives the surrounding environment based on laser or visible light; the positioning module is positioned based on GNSS (Global Navigation Satellite System, including Global Positioning System GPS, Beidou Positioning System BDS, Galileo satellite navigation system Galileo, and GLONASS) to obtain measured coordinate data; the processing module processes data; the storage module stores data generated during data processing and other data, for example, when installing the roadside device 100, the accurate coordinate data of the installation position of the roadside device 100 is determined by accurate surveying means, and the accurate coordinate data is stored in the storage module of the roadside device 100 and read out for use when needed; the communication module communicates with the cloud server through a wireless communication protocol.

[0056] In this embodiment, when steps S1-S4 are performed, step S1 can be performed by each roadside device to obtain first perception data, and the first perception data is sent to the cloud server; step S2 is performed by the vehicle terminal device to obtain second perception data, and the second perception data is sent to the cloud server; steps S3-S4 are performed by the cloud server to process the first perception data and the second perception data. The original data of each roadside device and vehicle terminal device can also be sent to the cloud server, and the cloud server can perform steps S1-S2 to obtain first perception data and second perception data, and perform steps S3-S4 to process the first perception data and the second perception data.

[0057] The flow of steps S1-S4 is as shown in Figure 3

[0058] In this embodiment, with reference to Figure 3 When step S1 is performed, that is, the road assets in the environment at the installation position are perceived by the plurality of roadside devices to obtain a plurality of first perception data, the following steps can be performed:

[0059] S101. For any roadside device, the first image data is obtained by performing point cloud and image perception scanning on the environment at the installation position by the roadside device;

[0060] S102. The first device coordinate data is obtained by positioning by the roadside device;

[0061] ​S103. Processing the first image data using a three-dimensional target detection algorithm to obtain first attribute data and first relative coordinate data;

[0062] S104. Determining first absolute coordinate data according to the first device coordinate data and the first relative coordinate data;

[0063] S105. Generating first perception data according to the first attribute data and the first absolute coordinate data.

[0064] In this embodiment, each roadside device performs steps S101-S105.

[0065] Taking one of the roadside devices as an example, in step S101, the roadside device performs point cloud and image perception scanning on the environment at the installation location through a point cloud and image perception scanning module to obtain first image data. The first image data reflects the information of the environment where the roadside device is located through images.

[0066] In step S102, the roadside device performs real-time positioning through a positioning module to obtain first device coordinate data. The first device coordinate data represents the coordinate information measured by the roadside device in real time.

[0067] In the case where the processing module of the roadside device has sufficient computing power, step S103 can be performed by the processing module of the roadside device, otherwise the first image data can be uploaded to a cloud server by the communication module of the roadside device, and step S103 can be performed by the cloud server. In step S103, the three-dimensional target detection algorithm is executed to process the first image data, so as to identify the first attribute data (which can represent the type of road asset, such as a bridge, a road sign, or a guardrail, and the number of the road asset, etc.) and the first relative coordinate data (which can represent the distance and angle of the road asset relative to the first device coordinate data, etc.) of the road asset (such as a bridge, a road sign, or a guardrail, etc.) contained in the first image data.

[0068] In step S104, the first relative coordinate data is mapped to the spatial coordinate system where the first device coordinate data is located through a spatial coordinate algorithm, so as to obtain first absolute coordinate data. The first absolute coordinate data can represent the absolute coordinate of the road asset in the first image data.

[0069] In step S105, the first attribute data and the first absolute coordinate data are packaged to obtain first perception data. Since the first perception data includes the first attribute data and the first absolute coordinate data, the first perception data can represent the coordinate and attribute of the road asset perceived by the roadside device.

[0070] In this embodiment, reference is made to Figure 3In step S2, that is, in the step of obtaining the second perception data by the vehicle-side device perceiving the road assets of the environment along the driving path, the following steps can be performed:

[0071] S201. Determine the coverage range of each roadside device according to the first device coordinate data of each roadside device.

[0072] S202. Plan the driving path of the vehicle-side device; the driving path bypasses the coverage range of each roadside device.

[0073] S203. Positioning by the vehicle-side device to obtain second device coordinate data.

[0074] S204. Driving along the driving path by the vehicle-side device to scan the environment along the driving path by point cloud and image perception to obtain second image data.

[0075] S205. Processing the second image data using a three-dimensional target detection algorithm to obtain second attribute data and second relative coordinate data; the second attribute data is used to represent the attributes of the road assets in the second image data, and the second relative coordinate data is used to represent the coordinates of the road assets in the second image data relative to the vehicle-side device.

[0076] S206. Determine the second absolute coordinate data according to the second device coordinate data and the second relative coordinate data; the second absolute coordinate data is used to represent the absolute coordinates of the road assets in the second image data.

[0077] S207. Generate the second perception data according to the second attribute data and the second absolute coordinate data.

[0078] In step S201, for any roadside device, a certain radius is set according to the performance of the roadside device with the first device coordinate data of the roadside device as the center, thereby determining the coverage range of the roadside device. Within the coverage range of the roadside device, the roadside device can scan to obtain qualified first image data, and outside the coverage range of the roadside device, it can be considered that the roadside device cannot scan to obtain qualified first image data.

[0079] In step S202, the driving path of the vehicle-side device is planned, and the planned driving path bypasses the coverage range of each roadside device, that is, when the vehicle drives along the driving path, the vehicle-side device will pass outside the coverage range of each roadside device, thereby enabling the vehicle-side device to detect road assets that cannot be detected by each roadside device.

[0080] The principles of steps S203-S207 are the same as those of steps S101-S105.

[0081] In step S203, the vehicle-side device obtains second device coordinate data by performing real-time positioning through the positioning module. The second device coordinate data represents the coordinate information measured by the vehicle-side device in real time.

[0082] In step S204, the vehicle-side device performs point cloud and image perception scanning on the environment at the installation location through the point cloud and image perception scanning module, and obtains second image data. The second image data reflects the information of the environment through which the vehicle-side device travels through images.

[0083] When the processing module of the vehicle-side device has sufficient computing power, step S205 can be performed by the processing module of the vehicle-side device. Otherwise, the vehicle-side device can upload the second image data to the cloud server through the communication module, and the cloud server can perform step S205. In step S205, a three-dimensional target detection algorithm is executed to process the second image data, so as to identify the second attribute data (which can represent that the road asset belongs to a bridge, a signpost, or a guardrail, and information such as the number of the road asset) and the second relative coordinate data (which can represent the distance and angle of the road asset relative to the vehicle-side device, i.e., the second device coordinate data) of the road asset (such as a bridge, a signpost, or a guardrail) contained in the second image data.

[0084] In step S206, the second relative coordinate data is mapped to the spatial coordinate system in which the second device coordinate data is located through a spatial coordinate algorithm, so as to obtain second absolute coordinate data. The second absolute coordinate data can represent the absolute coordinates of the road asset in the second image data.

[0085] In step S207, the second attribute data and the second absolute coordinate data are packaged to obtain second perception data. Since the second perception data includes the second attribute data and the second absolute coordinate data, the second perception data can represent the coordinates and attributes of the road asset perceived by the vehicle-side device.

[0086] In this embodiment, the following steps are performed with reference to Figure 3 When step S3 is performed, that is, when the first perception data and the second perception data are cooperatively optimized, the following steps can be performed:

[0087] S301. Each roadside device sets corresponding accurate coordinate data;

[0088] S302. Each roadside device determines difference correction data according to the corresponding accurate coordinate data and the first device coordinate data obtained by positioning;

[0089] S303. The second perception data is differentially corrected according to the difference correction data.

[0090] Steps S301-S303 are steps of differential positioning.

[0091] In step S301, when installing each roadside device 100, the precise coordinate data of the installation location of the roadside device 100 can be determined by precise surveying methods (e.g., confirming that the measurement environment is good and the measurement equipment is accurate enough). The precise coordinate data is stored in the storage module of the roadside device 100 and can be retrieved for use when needed.

[0092] Step S301 can be completed in one step when installing each roadside device. That is, if the storage module of the roadside device has already stored the precise coordinate data, the precise coordinate data can be read directly from the storage module.

[0093] In step S302, for the i-th roadside device, which serves as a reference station, the position of the navigation satellite can be calculated based on the ephemeris. Therefore, based on the position and precise coordinate data of the navigation satellite, the true geometric distance r between the i-th roadside device and the navigation satellite can be calculated. r (i) Based on the position of the navigation satellite and the measured coordinates of the first device, the pseudorange measurement value between the i-th roadside device and the navigation satellite can be calculated. Thus, it is possible to use the formula Calculate the differential correction data for the i-th roadside device. Among them, differential correction data This reflects errors caused by factors such as satellite clock bias, satellite ephemeris error, ionospheric delay, and tropospheric delay. The vehicle-mounted equipment can use differential correction data to correct these errors. (When there are a total of n roadside devices, i = 1, 2, ... n) to perform compensation, thereby eliminating or reducing these errors.

[0094] When performing step S303, the vehicle-mounted device activates the roadside device when it passes near the roadside device, and the roadside device sends the differential correction data calculated by the roadside device to the vehicle-mounted device. Alternatively, each roadside device can first upload its own differential correction data to the cloud server, and the cloud server can then forward all the differential correction data to the vehicle-mounted device.

[0095] When executing step S303, the vehicle-side equipment uses differential correction data. (i = 1, 2, ..., n) Compensate the second sensing data, thereby performing differential correction on the second sensing data, thus eliminating or reducing the error of the second device coordinate data in the second sensing data.

[0096] In this embodiment, when executing step S303, the following steps can be specifically performed:

[0097] S30301. Determine the environmental similarity between the vehicle-side equipment and each roadside equipment;

[0098] S30302. For any one differential correction data, according to the environment similarity between the corresponding roadside device and the vehicle-side device of the differential correction data, a corresponding weight is given to the differential correction data;

[0099] S30303. With each differential correction data, the second perception data is differentially corrected according to the respective corresponding weight.

[0100] In step S30301, the vehicle-side device and each roadside device detects the environmental parameters of the position where each device is located. The environmental parameters detected by each device are vectors of the same format, which can include one component or multiple components. For example, the environmental parameters can include two components of "longitude and latitude", or three components of "longitude, latitude and altitude", or multiple components such as "longitude, latitude, altitude, and weather condition quantization value (e.g. temperature, humidity, or discrete numerical values representing sunny, rainy, foggy weather, etc.)".

[0101] In step S30301, when determining the environment similarity between the vehicle-side device and the i-th roadside device, a vector similarity algorithm can be used to calculate the vector similarity between the environmental parameters of the vehicle-side device and the environmental parameters of the i-th roadside device as the environment similarity a i When only using two components of "longitude and latitude" or three components of "longitude, latitude and altitude" as environmental parameters, the inverse of the Euclidean distance between the two environmental parameters can be calculated as the environmental similarity, so the farther the vehicle-side device and the i-th roadside device, the smaller the environmental similarity a i , and the closer the environmental similarity a i is larger.

[0102] In step S30302, the differential correction data of the i-th roadside device According to the environment similarity a i between the corresponding roadside device and the vehicle-side device of the differential correction data, a corresponding weight is given to the differential correction data , that is, the differential correction data of the i-th roadside device after weighting is

[0103] In step S30303, the vehicle-side device obtains the differential correction data of all n roadside devices (i=1, 2, ……n), and the vehicle-side device compensates the second perception data according to the differential correction data (i=1, 2, ……n), thereby differentially correcting the second perception data, thereby eliminating or reducing the error of the second device coordinate data in the second perception data.

[0104] In steps S30301-S30303, the environmental parameters of the vehicle terminal device and each roadside device can reflect the environmental conditions of the locations where the devices are located, and the environmental similarity between the vehicle terminal device and each roadside device can reflect the degree of similarity of the environmental conditions of the locations where the devices are located. Since the positioning effect of the GNSS is related to the environmental conditions of the locations where the devices are located, the environmental similarity can be used as the GNSS positioning effect similarity of the vehicle terminal device and the ith roadside device. When the environmental similarity a i is smaller, the influence of the differential correction data provided by the ith roadside device on the differential correction of the vehicle terminal device is smaller, thereby facilitating further reduction of the errors in the second perception data caused by factors such as signal delay caused by weather.

[0105] In this embodiment, after step S3 is performed, that is, after the first perception data and the second perception data are cooperatively optimized, before step S4 is performed, that is, before the first perception data and the second perception data are integrated to obtain the asset information table, the first perception data and the second perception data are denoised. Specifically, denoising the first perception data and the second perception data includes the following steps:

[0106] P1. Establish a convolutional neural network;

[0107] P2. Obtain a first two-dimensional image;

[0108] P3. Add zero-mean noise to the first two-dimensional image to obtain a second two-dimensional image;

[0109] P4. Train the convolutional neural network by taking the second two-dimensional image as the input of the convolutional neural network and taking the first two-dimensional image as the expected output of the convolutional neural network;

[0110] P5. Convert the first perception data and the second perception data to obtain a third two-dimensional image;

[0111] P6. Input the third two-dimensional image into the trained convolutional neural network for processing.

[0112] The principles of steps P1-P6 are shown in Figure 4 The noise in the first perception data and the second perception data mainly comes from the pulse noise of the sensor, the jitter noise generated by the roadside device due to the passing of the vehicle, and the noise caused by the shaking of the vehicle terminal device when passing through different roads, etc. These noises are approximately close to zero-mean Gaussian noise, so the Noise2Noise denoising algorithm can be performed as shown in Figure 4

[0113] ​​Steps P1-P4 are the process of training the convolutional neural network.

[0114] In step P1, a convolutional neural network based on U-Net can be established. In step P2, the first two-dimensional image obtained is an image without zero-mean noise. In step P3, zero-mean noise is added to the first two-dimensional image to obtain a second two-dimensional image, so the content of the second two-dimensional image is the same as that of the first two-dimensional image, and the second two-dimensional image contains zero-mean noise compared with the first two-dimensional image.

[0115] In step P4, the second two-dimensional image is taken as the input of the convolutional neural network, and the first two-dimensional image is taken as the expected output (label) of the convolutional neural network. The actual output obtained by processing the second two-dimensional image by the convolutional neural network is compared with the label to calculate the loss function value using the L2 norm. When the loss function value does not converge, the network parameters of the convolutional neural network are updated, and the next set of first two-dimensional image and second two-dimensional image are read for continuous training. When the loss function value does not converge, the training of the convolutional neural network is ended.

[0116] After the training of steps P1-P4, the obtained convolutional neural network has the ability to process two-dimensional images and eliminate zero-mean noise in them.

[0117] In step P5, the first perception data and the second perception data are respectively converted to obtain a third two-dimensional image, so that the first perception data and the second perception data are converted into a data format that can be processed by the convolutional neural network. In step P6, the third two-dimensional image is input into the trained convolutional neural network for processing, so that the convolutional neural network can eliminate the impulse noise and jitter noise in the first perception data and the second perception data, thereby obtaining more accurate results.

[0118] After steps P1-P5 are performed, step S5 is performed to integrate the first perception data and the second perception data with the noise eliminated to obtain an asset information table.

[0119] In step S5, the first attribute data in the first perception data labels the attributes (including the name, number, type, etc.) of part of the road assets, and the first absolute coordinate data labels the positions of the part of the road assets. The second attribute data in the second perception data also labels the attributes of part of the road assets, and the second absolute coordinate data labels the positions of the part of the road assets. Therefore, the following situations can exist:

[0120] (1) Some road assets labeled in the first perception data are not labeled in the second perception data;

[0121] (2) Some road assets labeled in the second perception data are not labeled in the first perception data;

[0122] (3) some road assets calibrated in the first perception data have been calibrated in the second perception data, and there is no or little deviation between the first absolute coordinate data and the second absolute coordinate data corresponding to the same road asset;

[0123] (4) some road assets calibrated in the first perception data have been calibrated in the second perception data, but there is large deviation between the first absolute coordinate data and the second absolute coordinate data corresponding to the same road asset.

[0124] For the road assets in the first and second cases, the attribute data and absolute coordinate data thereof are recorded in the asset information table; for the road assets in the third case, the attribute data and absolute coordinate data thereof are recorded in the asset information table after deduplication; and for the road assets in the fourth case, the same are marked as ambiguous assets in the asset information table.

[0125] In the embodiment, the road asset detection management method based on vehicle-road cooperation further performs the following steps:

[0126] S5. Comparing the asset information tables obtained respectively by performing multiple detection cycles;

[0127] S6. Performing asset management according to the comparison result.

[0128] Steps S1-S4 are the results obtained in one detection cycle. By performing multiple detection cycles, one asset information table can be obtained in each detection cycle, thereby obtaining multiple asset information tables. When performing step S5, two asset information tables obtained in two adjacent detection cycles can be compared, thereby determining the changes of road assets in the asset information table, such as addition and deletion. In step S6, when the comparison result of the two asset information tables obtained in two adjacent detection cycles shows that a road asset is added (a set of completely new attribute data and absolute coordinate data is recorded in the latter asset information table) and deleted (a set of attribute data and absolute coordinate data appears in the former asset information table, but no longer appears in the former asset information table), the added or removed road asset is checked and managed; when only the attribute data of a road asset changes without change in the absolute coordinate data, or only the absolute coordinate data of the road asset changes without change in the attribute data, etc., the asset can be marked as an ambiguous asset, thereby judging whether it is damaged, stolen, and performing manual repair and other operations.

[0129] The vehicle-road cooperation based road asset detection management method in the embodiment can be implemented by writing a computer program for executing the vehicle-road cooperation based road asset detection management method, writing the computer program into a storage medium or a computer device, and executing the vehicle-road cooperation based road asset detection management method when the computer program is read out and run, so as to achieve the same technical effects as the vehicle-road cooperation based road asset detection management method in the embodiment.

[0130] It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed", "connected" to another feature, it can be directly fixed, connected to the other feature, or indirectly fixed, connected to the other feature. In addition, the up, down, left, right and the like used in the disclosure are only relative to the relative positional relationship of the components of the disclosure in the drawings. The singular forms "a", "an" and "the" used in the disclosure are also intended to include the plural forms, unless the context clearly indicates otherwise. In addition, unless otherwise defined, all technical and scientific terms used in the embodiments have the same meaning as generally understood by those skilled in the art. The terms used in the embodiments of the present disclosure are only used to describe the specific embodiments, and are not intended to limit the present disclosure. The term "and / or" used in the embodiments includes any combination of one or more related listed items.

[0131] It should be understood that although the terms first, second, third, etc. can be used in the present disclosure to describe various elements, these elements should not be limited to these terms. These terms are only used to distinguish one type of element from another. For example, without departing from the scope of the present disclosure, a first element can also be referred to as a second element, and similarly, a second element can also be referred to as a first element. The use of any and all examples or exemplary language (e.g., "for example", "as such", etc.) provided in the embodiments of the present disclosure is only intended to better illustrate the embodiments of the present disclosure, and does not impose any limitation on the scope of the present disclosure unless otherwise required.

[0132] It should be recognized that the embodiments of the present disclosure can be implemented or embodied by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The method can be implemented in a computer program configured with a non-transitory computer readable storage medium, using standard programming techniques - including the non-transitory computer readable storage medium configured with the computer program, in which the storage medium so configured causes the computer to operate in a specific and predefined manner - according to the methods described in the specific embodiments and the drawings. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, if necessary, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can be run on a programmed special-purpose integrated circuit for this purpose.

[0133] Further, the operations of the processes described in this embodiment can be performed in any suitable order, unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described in this embodiment (or variations and / or combinations thereof) can be implemented under the control of one or more computer systems configured with executable instructions (e.g., computer programs, one or more computer processes, or one or more applications) that can be executed by one or more processors, by hardware, or by a combination thereof. The computer programs include a plurality of instructions that are executable by one or more processors.

[0134] Further, the methods can be implemented in any suitable type of computing platform operatively coupled to, including but not limited to, a personal computer, a mini-computer, a mainframe, a workstation, a network or distributed computing environment, a stand-alone or integrated computer platform, or in communication with a charged particle tool or other imaging device, and the like. Aspects of the present application can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated to the computing platform, such as a hard disk, an optical read and / or write storage medium, RAM, ROM, and the like, such that it can be read by a programmable computer to configure and operate the computer to perform the processes described herein when the storage medium or device is read by the computer. Further, the machine-readable code, or portions thereof, can be transmitted over wired or wireless networks. The present application described in this embodiment includes these and other different types of non-transitory computer readable storage media when such media include instructions or programs to implement the steps described above in conjunction with a microprocessor or other data processor. The present application also includes the computer itself when programmed according to the methods and techniques described in the present application.

[0135] The computer programs are capable of applying to input data to perform the functions described in this embodiment, thereby transforming the input data to generate output data that is stored to non-volatile memory. The output information can also be applied to one or more output devices, such as a display. In preferred embodiments of the present application, the transformed data represents a physical and tangible object, including a particular visual depiction of the physical and tangible object produced on a display.

[0136] The above description is only preferred embodiments of the present application, the present application is not limited to the above-described embodiments, as long as the same means to achieve the technical effects of the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application. The technical solutions and / or embodiments within the scope of protection of the present application can have various modifications and changes.

Claims

1. A vehicle infrastructure integration based road asset detection management method, characterized in that, The vehicle-road cooperation-based road asset detection management method performs a plurality of detection cycles, and in each detection cycle, the following steps are performed: A plurality of roadside devices perceive road assets in the environment where the installation position is located, and obtain a plurality of first perception data; the first perception data are used to represent the coordinates and attributes of the road assets perceived by the roadside devices; A vehicle-side device perceives road assets in the environment through which a driving path passes, and obtains second perception data; the second perception data are used to represent the coordinates and attributes of the road assets perceived by the vehicle-side device; The first perception data and the second perception data are cooperatively optimized; The first perception data and the second perception data are integrated to obtain an asset information table; The cooperative optimization of the first perception data and the second perception data includes: Each of the roadside devices sets corresponding accurate coordinate data; the accurate coordinate data are used to represent the coordinates of the installation position of the corresponding roadside device; Each of the roadside devices determines difference correction data according to the corresponding accurate coordinate data and the first device coordinate data obtained by positioning; The vehicle-side device and each of the roadside devices detect environmental parameters at their respective positions, calculate the vector similarity between the environmental parameters of the vehicle-side device and the environmental parameters of each of the roadside devices, and determine the environmental similarity between the vehicle-side device and each of the roadside devices; For any one of the difference correction data, a corresponding weight is given to the difference correction data according to the environmental similarity between the corresponding roadside device and the vehicle-side device of the difference correction data; The second perception data are differentially corrected according to each of the difference correction data and the corresponding weight. 2.The CAV-based road asset detection management method of claim 1, wherein, The plurality of roadside devices perceive road assets in the environment where the installation position is located, and obtain a plurality of first perception data, including: For any one of the roadside devices, the roadside device performs point cloud and image perception scanning on the environment where the installation position is located, and obtains first image data; The roadside device is positioned to obtain first device coordinate data; The first image data are processed using a three-dimensional target detection algorithm to obtain first attribute data and first relative coordinate data; the first attribute data are used to represent the attributes of the road assets in the first image data, and the first relative coordinate data are used to represent the coordinates of the road assets in the first image data relative to the roadside device; First absolute coordinate data are determined according to the first device coordinate data and the first relative coordinate data; the first absolute coordinate data are used to represent the absolute coordinates of the road assets in the first image data; The first attribute data and the first absolute coordinate data are used to generate the first perception data. 3.The CAV-based road asset detection management method of claim 2, wherein, The vehicle-side device perceives road assets in the environment through which a driving path passes, and obtains second perception data, including: According to each of the first device coordinate data, the coverage range of each of the roadside devices is determined; planning a driving path of the vehicle-side device; the driving path bypasses the coverage range of each roadside device; obtaining second device coordinate data by positioning through the vehicle-side device; obtaining second image data by driving along the driving path through the vehicle-side device, and performing point cloud and image perception scanning on the environment passed through by the driving path; processing the second image data using a three-dimensional target detection algorithm to obtain second attribute data and second relative coordinate data; the second attribute data is used to represent the attributes of the road assets in the second image data, and the second relative coordinate data is used to represent the coordinates of the road assets in the second image data relative to the vehicle-side device; determining second absolute coordinate data according to the second device coordinate data and the second relative coordinate data; the second absolute coordinate data is used to represent the absolute coordinates of the road assets in the second image data; generating the second perception data according to the second attribute data and the second absolute coordinate data. 4.The CAV-based road asset detection management method of claim 1, wherein, In any of the detection cycles, the following steps are also performed: After the first perception data and the second perception data are cooperatively optimized, the first perception data and the second perception data are denoised before the asset information table is obtained by integrating the first perception data and the second perception data. 5.The CAV-based road asset detection management method of claim 4, wherein, The denoising of the first perception data and the second perception data includes: establishing a convolutional neural network; obtaining a first two-dimensional image; adding zero-mean noise to the first two-dimensional image to obtain a second two-dimensional image; training the convolutional neural network by taking the second two-dimensional image as the input of the convolutional neural network and taking the first two-dimensional image as the expected output of the convolutional neural network; converting the first perception data and the second perception data to obtain a third two-dimensional image; inputting the third two-dimensional image into the trained convolutional neural network for processing. 6.The CAV-based road asset detection management method according to any one of claims 1-5, characterized in that, The vehicle-road cooperation-based road asset detection management method further includes the following steps: comparing the asset information tables obtained by performing a plurality of detection cycles respectively; performing asset management according to the comparison result.

7. A computer apparatus, comprising: The memory is used to store at least one program, and the processor is used to load the at least one program to execute the vehicle-road cooperation-based road asset detection management method according to any one of claims 1-6.

8. A computer readable storage medium having stored therein a program which is executable by a processor, characterized in that, The program executable by the processor, when executed by the processor, is used to execute the vehicle-road cooperation-based road asset detection management method according to any one of claims 1-6.

Images