Data processing method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a data processing method. The method comprises the following steps: acquiring a target disease data set; performing traversal calculation in the target disease data set based on the recognition time and the disease position according to the sequence of the recognition time to obtain a disease distance between the first road disease and the second road disease, a recognition time interval and a driving distance of the acquisition device; determining whether to add the first road disease and the second road disease to a similar disease data set based on the disease distance, the identification time interval and the driving distance; and after traversal calculation is completed, selecting target road disease data in the similar disease data set. Road disease data of the same type are collected into one data set for unified processing, the data processing efficiency can be improved, the accurate similar data set can be obtained based on the fact that multi-dimensional data serve as the basis of whether the road diseases are included in the similar data set or not, and therefore after traversal is completed, the data processing efficiency is improved. The target road disease data can be accurately selected from the similar data set.

Description

Technical Field

[0001] The present invention relates to the field of equipment, and in particular to a data processing method, device, electronic equipment and storage medium. Background Art

[0002] In the traditional road inspection process, the road information is usually collected by the collection device in a mobile manner, and the road defects are analyzed and captured, and then the defects are uploaded to the road inspection system. In the process of collecting road information by the collection device, due to the change in the relative position of the collection device and the road surface defects, the collection device will repeatedly collect the same defect and report it to the inspection system, resulting in a large amount of repeated and redundant disease information being collected, which leads to the inspection system being less efficient when performing functions such as analysis and statistics. At present, a more feasible way to solve the above problem is to remove duplicates through manual review, but manual review is subjective, its deduplication accuracy is not high, and its efficiency is still not enough to meet most usage scenarios. Therefore, how to provide a data processing method that can improve the data processing accuracy and data processing efficiency when eliminating redundant disease information has become a problem that needs to be solved urgently. Summary of the invention

[0003] The embodiment of the present invention provides a data processing method, which aims to solve the problem of low data processing accuracy and data processing efficiency when the existing data processing method removes redundant disease information. By combining the same type of road disease data into one data set for unified processing, the data processing efficiency can be improved, and in the order of recognition time, based on the recognition time and the disease location, the multi-dimensional data such as the disease distance between road diseases, the recognition time interval and the driving distance of the collection device are traversed and calculated in the data set, and then based on the multi-dimensional data as the basis for whether the road disease is included in the similar data set, an accurate similar data set can be obtained, so that after the traversal is completed, the target road disease data can be accurately selected in the similar data set for analysis and statistics and other functions.

[0004] In a first aspect, an embodiment of the present invention provides a data processing method, the method comprising the following steps:

[0005] Acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease;

[0006] In the order of the recognition time, based on the recognition time and the disease location, the target disease data set is traversed and calculated to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease;

[0007] Determining whether to add the first road disease and the second road disease to a similar disease dataset based on the disease distance, the identification time interval, and the driving distance;

[0008] After the traversal calculation is completed, target road disease data is selected from the similar disease data set.

[0009] Optionally, the step of traversing and calculating in the target disease data set in the order of the recognition time and based on the recognition time and the disease location to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device includes:

[0010] Extracting road defects from the target defect data set one by one as the first road defects in the order of the identification time;

[0011] Whenever one of the first road diseases is extracted, a road disease in the target disease data set whose identification time is later than that of the first road disease is taken as the second road disease;

[0012] The second road defects are selected one by one from the target data set in the order of the recognition time, and based on the recognition time and the defect location, the defect distance between the second road defect and the first road defect, the recognition time interval and the travel distance of the acquisition device are calculated.

[0013] Optionally, the road damage data includes the driving speed when the collection device identifies the road damage, and the calculation of the damage distance between the second road damage and the first road damage, the identification time interval, and the driving distance of the collection device based on the identification time and the damage location includes:

[0014] Based on the recognition time, calculating the recognition time interval between the second road defect and the first road defect;

[0015] Based on the defect location, calculating the defect distance between the second road defect and the first road defect;

[0016] Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0017] The driving distance is calculated based on the average driving speed and the identification time interval.

[0018] Optionally, the road damage data includes the driving speed when the acquisition device identifies the road damage, and determining whether to add the first road damage and the second road damage to the similar damage data set based on the damage distance, the identification time interval and the driving distance includes:

[0019] Based on the type of road damage in the target damage data set, a speed threshold and a distance threshold are set;

[0020] Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0021] Determining a time threshold corresponding to the first road defect and the second road defect based on the average driving speed and the speed threshold;

[0022] Based on the disease distance and the travel distance, calculating the remaining distance interval corresponding to the first road disease and the second road disease;

[0023] When the identification time interval corresponding to the first road defect and the second road defect is less than the time threshold, and the corresponding remaining distance interval is less than the distance threshold, the first road defect and the second road defect are added to the similar defect data set.

[0024] Optionally, the road damage data includes recognition confidence of the road damage, and the selecting target road damage data from the similar damage data set includes:

[0025] Extracting the road disease data with the highest recognition confidence from the similar disease data set;

[0026] When the number of the road damage data with the highest recognition confidence is one, determining the road damage data with the highest recognition confidence as the target road damage data;

[0027] When there are multiple road damage data with the highest recognition confidence, the target road damage data is determined according to the recognition time of the multiple road damage data with the highest recognition confidence.

[0028] Optionally, before obtaining the target disease dataset, the method further includes:

[0029] Acquire multiple disease images of the same type, image coordinates of road diseases in the disease images, and acquisition parameters of the target disease image;

[0030] Normalizing the size of the target disease image to obtain a target disease image;

[0031] Determining the location of the road disease in the target disease image based on the image coordinates and the acquisition parameters;

[0032] Based on the target disease image and the disease location, the target disease dataset is constructed.

[0033] Optionally, the acquisition parameters include an acquisition height and an internal reference matrix, and determining the defect position of the road defect in the target defect image based on the image coordinates and the acquisition parameters includes:

[0034] Determining the three-dimensional coordinates of the road damage based on the image coordinates and the internal parameter matrix;

[0035] Determining the dedistorted three-dimensional coordinates of the road defect based on the three-dimensional coordinates and a preset dedistortion coefficient;

[0036] The defect location of the road defect is determined based on the dedistorted three-dimensional coordinates of the road defect and the acquisition height.

[0037] In a second aspect, an embodiment of the present invention further provides a data processing device, the data processing device comprising:

[0038] A first acquisition module is used to acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease;

[0039] A first calculation module is used to traverse and calculate in the target disease data set in the order of the recognition time, based on the recognition time and the disease location, to obtain the disease distance between the first road disease and the second road disease, the recognition time interval and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease;

[0040] A first determination module, configured to determine whether to add the first road disease and the second road disease to a similar disease data set based on the disease distance, the identification time interval, and the driving distance;

[0041] The first selection module is used to select target road disease data from the similar disease data set after the traversal calculation is completed.

[0042] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the data processing method provided in the embodiment of the present invention when executing the computer program.

[0043] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the data processing method provided in the embodiment of the invention are implemented.

[0044] In an embodiment of the present invention, a target disease data set is obtained, the target disease data set includes multiple road disease data of the same type, the road disease data includes the recognition time of the road disease and the disease position of the road disease, the recognition time is the time when the acquisition device recognizes the road disease; in the order of the recognition time, based on the recognition time and the disease position, the target disease data set is traversed and calculated to obtain the disease distance, recognition time interval and driving distance between the first road disease and the second road disease, the recognition time of the first road disease is earlier than the recognition time of the second road disease; based on the disease distance, the recognition time interval and the driving distance, it is determined whether to add the first road disease and the second road disease to the similar disease data set; after the traversal calculation is completed, the target road disease data is selected from the similar disease data set. By combining the same type of road disease data into one data set for unified processing, the data processing efficiency can be improved. In the order of recognition time, based on the recognition time and the location of the disease, the multi-dimensional data such as the disease distance between road diseases, the recognition time interval and the driving distance of the collection device are traversed and calculated in the data set. Based on the multi-dimensional data as the basis for whether the road disease is included in the similar data set, an accurate similar data set can be obtained. Therefore, after the traversal is completed, the target road disease data can be accurately selected from the similar data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 is a flow chart of a data processing method provided by an embodiment of the present invention;

[0047] Figure 2 is a flow chart of another data processing method provided by an embodiment of the present invention;

[0048] Figure 3 is a structural schematic diagram of a data processing device provided in an embodiment of the present invention;

[0049] Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] like Figure 1 As shown, Figure 1 is a flow chart of a data processing method provided by an embodiment of the present invention, comprising:

[0052] 101. Obtain target disease dataset.

[0053] In an embodiment of the present invention, the above-mentioned data processing method can be applied to a road inspection system, the above-mentioned road inspection system can be mounted on a road inspection platform, the above-mentioned road inspection platform can be constructed through a server or a server platform, and the above-mentioned server or server platform can be any electronic device with functions such as data processing, data analysis, data storage, and data transmission.

[0054] The target disease data set includes multiple road disease data of the same type, and the road disease data includes the recognition time of the road disease and the disease location of the road disease. The recognition time is the time when the acquisition device recognizes the road disease. The type can be understood as the disease type, and the disease can be understood as various problems and damages that affect the service life of the road and driving safety. For example, it can be road cracks, road potholes, road fractures, road subsidence, etc. The disease location can be the geographical coordinates of the road disease, for example, it can be the longitude and latitude coordinates.

[0055] When the acquisition device (or road inspection device) is inspecting on the road, it will collect road images through the image acquisition device deployed on the acquisition device, and identify the defects of the road images. When a road defect is identified, the current time (i.e., the identification time), the longitude and latitude of the acquisition device, and the speed will be recorded. Then, the type and confidence of the defect are identified, and the location of the defect in the picture (i.e., the image coordinates) is marked, and a unique ID label is assigned to the road defect.

[0056] After assigning labels to road defects, the collection device can calculate the damage location of the road defect based on the above-mentioned image coordinates, and collect the damage location, identification time, label and corresponding damage image (i.e., road image containing road defects) of the road defect into a road defect data.

[0057] The acquisition frequency of the acquisition device can be determined according to the frame rate of the video acquired by the image acquisition device. For example, a 30-frame video means that 30 times (i.e., 30 road damage data) will be acquired in 1 second. The acquisition device will collect the road damage data of the same damage type acquired within a preset time as a damage data set and upload it to the road inspection platform. The preset time can be set according to the storage space of the acquisition device and the processing performance of the road inspection system. The smaller the storage space and the weaker the processing performance, the shorter the preset time can be set accordingly. The larger the storage space and the stronger the processing performance, the longer the preset time can be set accordingly.

[0058] After receiving (or understanding as obtaining) the above-mentioned disease data set, the above-mentioned road inspection platform can directly determine the above-mentioned disease data set as the above-mentioned target disease data set, or when the processing performance of the road inspection system is sufficient, after uploading multiple disease data sets on the same acquisition device, when the data scale of the multiple disease data sets meets the preset scale threshold, the multiple disease data sets can be combined into the target disease data set. The above-mentioned preset scale threshold can also be set according to the processing performance of the road inspection system. When the above-mentioned processing performance is stronger, the above-mentioned preset scale threshold can be set larger accordingly, and when the above-mentioned processing performance is weaker, the above-mentioned preset scale threshold can be set smaller accordingly.

[0059] Alternatively, the above-mentioned collection device can directly upload the road disease data to the above-mentioned road inspection platform. The above-mentioned road inspection platform can perform a first classification on the acquired road disease data according to the collection device to obtain the road disease data of the same collection device, perform a second classification on the road disease data of the same collection device according to the disease type to obtain the road disease data of the same collection device and the same type, and add the road disease data of the same collection device and the same type to the disease data set. When the data scale of any disease data set meets the above-mentioned preset scale threshold, the above-mentioned disease data set is determined as the above-mentioned target disease data set.

[0060] After acquiring the target disease data set, the road inspection platform may execute any one of the data processing methods provided in the embodiments of the present invention to determine the target disease data.

[0061] 102. In the order of recognition time, based on the recognition time and the location of the defect, traverse and calculate in the target defect data set to obtain the defect distance between the first road defect and the second road defect, the recognition time interval, and the travel distance of the collection device.

[0062] In the embodiment of the present invention, the above-mentioned disease distance may be the geographical distance between the center points of two road diseases, and the distance unit may be centimeters, decimeters, meters, etc., which may be set according to actual needs. The above-mentioned identification time interval may be the time difference between the identification times of the two road diseases, and the travel distance of the above-mentioned collection device may be understood as the distance that the collection device needs to move when collecting the above-mentioned second road disease after collecting the above-mentioned first road disease. The identification time of the above-mentioned first road disease is earlier than the identification time of the above-mentioned second road disease.

[0063] The above traversal calculation can be explained by the concept of calculation rounds. In each calculation round, the road disease corresponding to the current calculation round can be extracted from the target disease data set as the first road disease, and the road disease in the target disease data set whose identification time is later than the first road disease is determined as the second road disease. The disease distance, identification time interval and driving distance of the collection device between the first road disease and each second road disease are calculated one by one. After the disease distance, identification time interval and driving distance of the collection device between the first road disease and all the second road diseases are calculated in the current calculation round, the current calculation round is ended. Enter the next calculation round.

[0064] It should be noted that the road disease data in the above-mentioned target disease data set can be arranged according to the identification time corresponding to each road disease data. If the current calculation round is the first calculation round, then the first road disease corresponding to the current calculation round is the road disease with the earliest identification time in the target data set. If the current calculation round is the second calculation round, then the first road disease corresponding to the current calculation round is the first road disease in the target data set whose identification time is later than the first calculation round, but earlier than other road diseases. If the current calculation round is the third calculation round, then the first road disease corresponding to the current calculation round is the first road disease in the target data set whose identification time is later than the first calculation round and the first road disease in the second calculation round, but earlier than other road diseases. Until the road disease with the latest identification time is taken as the first road disease, since there are no other road diseases with identification time later than the first road disease, the traversal calculation can be terminated.

[0065] Specifically, in the above target disease data set, each road disease data may include the recognition time and the disease location of the road disease. The recognition time interval between each two road diseases can be calculated based on the recognition time, the disease distance between each two road diseases can be calculated based on the disease location, and the driving distance when the collection device collects each two road diseases can be calculated based on the recognition time interval.

[0066] 103. Determine whether to add the first road disease and the second road disease to a similar disease data set based on the disease distance, the identification time interval, and the driving distance.

[0067] In an embodiment of the present invention, a speed threshold and a distance threshold can be set according to the type of road damage in the target damage data set, and the average driving speed corresponding to the first road damage and the second road damage identified by the acquisition device can be calculated. According to the average driving speed and the speed threshold, the time threshold corresponding to the first road damage and the second road damage can be determined. At the same time, according to the damage distance and driving distance between the first road damage and the second road damage, the remaining distance interval corresponding to the first road damage and the second road damage can be determined.

[0068] The identification time intervals corresponding to the first road defect and the second road defect are compared with the time threshold to obtain a first comparison result. The remaining distance intervals corresponding to the first road defect and the second road defect are compared with the distance threshold to obtain a second comparison result. According to the first comparison result and the second comparison result, it is determined whether to add the first road defect and the second road defect to the similar defect data set.

[0069] For example, the type of the road disease can be the above-mentioned disease type. Different disease types have different impact ranges on the road. When the above-mentioned disease type is road cracks, since the impact range of cracks on the road is usually small, when the collection device is patrolling on the above-mentioned road, the speed and driving distance will be slightly affected. Therefore, the above-mentioned speed threshold and distance threshold can be set to be larger accordingly. When the above-mentioned disease type is road subsidence, since the subsidence is usually large-scale subsidence, the impact range on the road is large. When the collection device is patrolling on the above-mentioned road, the speed and driving distance will be seriously affected. Therefore, the above-mentioned speed threshold and distance threshold can be set to be smaller accordingly. In summary, the above-mentioned speed threshold and distance threshold can be inversely proportional to the impact range corresponding to the disease type.

[0070] It should be noted that the speed threshold and distance threshold are set according to the type of road damage, while the time threshold is set according to the average driving speed corresponding to the two road damages. Therefore, the speed threshold, distance threshold and time threshold are set dynamically. When the type of road damage and the average driving speed change, the threshold will increase or decrease accordingly, which can more accurately determine whether the first road damage and the second road damage need to be added to the similar damage data set.

[0071] 104. After the traversal calculation is completed, the target road disease data is selected from the similar disease data set.

[0072] In the embodiment of the present invention, when the acquisition device identifies the above road damage in the road image, it will synchronously output the recognition confidence. Therefore, after the traversal calculation is completed, each road damage data in the similar damage data set can be sorted according to the recognition confidence, so as to determine the road damage data with the highest recognition confidence, and determine the road damage data with the highest recognition confidence as the target road damage data.

[0073] In an embodiment of the present invention, a target disease data set is obtained, the target disease data set includes multiple road disease data of the same type, the road disease data includes the recognition time of the road disease and the disease position of the road disease, the recognition time is the time when the acquisition device recognizes the road disease; in the order of the recognition time, based on the recognition time and the disease position, the target disease data set is traversed and calculated to obtain the disease distance, recognition time interval and driving distance between the first road disease and the second road disease, the recognition time of the first road disease is earlier than the recognition time of the second road disease; based on the disease distance, the recognition time interval and the driving distance, it is determined whether to add the first road disease and the second road disease to the similar disease data set; after the traversal calculation is completed, the target road disease data is selected from the similar disease data set. By combining the same type of road disease data into one data set for unified processing, the data processing efficiency can be improved. In the order of recognition time, based on the recognition time and the location of the disease, the multi-dimensional data such as the disease distance between road diseases, the recognition time interval and the driving distance of the collection device are traversed and calculated in the data set. Based on the multi-dimensional data as the basis for whether the road disease is included in the similar data set, an accurate similar data set can be obtained. Therefore, after the traversal is completed, the target road disease data can be accurately selected from the similar data set.

[0074] It is understandable that in the specific implementation of this application, related data such as disease data sets and disease images are involved. When the embodiments in this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data and the construction, training and use of related models need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0075] It should be noted that the data processing method provided in the embodiment of the present invention can be applied to devices such as computers and servers that can perform data processing.

[0076] Optionally, in the step of traversing and calculating in the target disease data set in order of recognition time, based on the recognition time and the disease location, to obtain the disease distance between the first road disease and the second road disease, the recognition time interval and the driving distance of the collection device, the road diseases can also be extracted one by one from the target disease data set in order of recognition time as the first road disease; each time a first road disease is extracted, the road disease in the target disease data set whose recognition time is later than that of the first road disease is taken as the second road disease; the second road diseases are selected one by one from the target data set in order of recognition time, and the disease distance between the second road disease and the first road disease, the recognition time interval and the driving distance of the collection device are calculated based on the recognition time and the disease location.

[0077] In the embodiment of the present invention, the order of recognition time may be from early to late. Assuming that the target disease data set includes road disease A0, road disease A1, road disease A2 and road disease A3, the recognition time of A0 is earlier than that of A1, the recognition time of A1 is earlier than that of A2, and the recognition time of A2 is earlier than that of A3.

[0078] To further explain the concept of calculation rounds, in the first calculation round, A0 is the first road disease, A1, A2, and A3 are the second road diseases. In the second calculation round, A1 is the first road disease, A2 and A3 are the second road diseases. In the third calculation round, A2 is the first road disease, and A3 is the second road disease. In the fourth calculation round, A3 is the first road disease. Since there is no road disease whose identification time is later than A3, the traversal calculation can be ended at this time.

[0079] In the first calculation round, A1 can be selected from the target disease data set in the order of recognition time, and the disease distance, recognition time interval and driving distance of the collection device between A0 and A1 are calculated. After the calculation is completed, A2 is selected, and the disease distance, recognition time interval and driving distance of the collection device between A0 and A2 are calculated. After the calculation is completed, A3 is selected, and the disease distance, recognition time interval and driving distance of the collection device between A0 and A3 are calculated. After the calculation is completed, the calculation of the first calculation round is terminated and the second calculation round is entered. The selection method of subsequent rounds is the same as the selection method of the first calculation round mentioned above, and the effect is the same. In order to avoid repetition and redundancy, it will not be repeated here.

[0080] Optionally, in the step of calculating the defect distance, identification time interval and driving distance of the collection device between the second road defect and the first road defect based on the identification time and the defect location, the identification time interval between the second road defect and the first road defect can also be calculated based on the identification time; the defect distance between the second road defect and the first road defect can be calculated based on the defect location; the average driving speed can be calculated based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect; and the driving distance can be calculated based on the average driving speed and the identification time interval.

[0081] In the embodiment of the present invention, the above-mentioned collection device often identifies road damage during the inspection process while the vehicle is in the process of driving, so the above-mentioned road damage data includes the driving speed when the collection device identifies the road damage. The above-mentioned identification time interval can be calculated by the following formula:

[0082] Xs=S1-S0

[0083] Among them, the above Xs is the identification time interval, the above S1 represents the identification time of the second road disease, and the above S0 represents the identification time of the first road disease.

[0084] The above-mentioned disease location can be the longitude and latitude of the road disease. The above-mentioned longitude and longitude can include the longitude and latitude of the road disease. The longitude difference between the center point of the first road disease and the center point of the second road disease can be calculated based on the longitude of the center point of the first road disease and the longitude of the center point of the second road disease. The latitude difference between the center point of the first road disease and the center point of the second road disease can be calculated based on the latitude of the center point of the first road disease and the latitude of the center point of the second road disease. The disease distance between the first road disease and the second road disease can be calculated based on the above-mentioned longitude difference, latitude difference and the radius of the earth.

[0085] Specifically, the above disease distance can be calculated by the following formula:

[0086]

[0087] Among them, the above x d is represented by the above disease distance, and the above r is represented by the radius of the earth. The above radius can be the radius of the earth in the WGS-84 standard, that is, 6378137 meters. Expressed as the above latitude difference (i.e. ), the above It is expressed as the latitude of the center point of the first road disease. It represents the latitude of the center point of the second road defect, the above Δλ represents the above longitude difference (ie Δλ=λ1-λ2), the above λ1 represents the longitude of the center point of the first road defect, and the above λ2 represents the longitude of the center point of the second road defect.

[0088] The above driving distance can be calculated by the following formula:

[0089]

[0090] Among them, the above x r Expressed as the above driving distance, the above It is represented by the above average distance, the above v0 is represented by the driving speed of the collection device when it identifies the above first road damage, the above v1 is represented by the driving speed of the collection device when it identifies the above second road damage, and the above Xs is represented by the above identification time interval.

[0091] Optionally, in the step of determining whether to add the first road disease and the second road disease to the similar disease data set based on the disease distance, the identification time interval and the driving distance, a speed threshold and a distance threshold can also be set based on the type of road diseases in the target disease data set; the average driving speed is calculated based on the driving speed corresponding to the first road disease and the driving speed corresponding to the second road disease; the time threshold corresponding to the first road disease and the second road disease is determined based on the average driving speed and the speed threshold; the remaining distance interval corresponding to the first road disease and the second road disease is calculated based on the disease distance and the driving distance; when the identification time interval corresponding to the first road disease and the second road disease is less than the time threshold, and the corresponding remaining distance interval is less than the distance threshold, the first road disease and the second road disease are added to the similar disease data set.

[0092] In an embodiment of the present invention, the road damage data includes the driving speed when the acquisition device identifies the road damage, the type of the road damage may be the damage type, the impact range of different damage types on the road may be determined, and the impact may be the impact on road traffic. According to the impact range, the speed threshold and the distance threshold are set. Specifically, the impact range is inversely proportional to the distance threshold and the speed threshold. When the impact range is larger, the distance threshold and the speed threshold can be set smaller accordingly, and when the impact range is smaller, the distance threshold and the speed threshold can be set larger accordingly.

[0093] After determining the speed threshold and the distance threshold, assuming that the speed threshold is k, the time threshold can be further described by the following formula:

[0094]

[0095] Among them, the above S represents the above time threshold, the above It is represented as the above average driving speed. The calculation method of the above average driving speed is consistent with the calculation method of the average driving speed in the above formula of driving distance. The above k is represented as the above speed threshold.

[0096] The remaining distance interval can be calculated by the following formula:

[0097] x rd =x d -x r

[0098] Among them, the above x rd Expressed as the above remaining distance interval, the above x d Expressed as the above disease distance, the above x r Expressed as the driving distance above.

[0099] Optionally, in the step of selecting target road damage data from a similar damage data set, road damage data with the highest recognition confidence can also be extracted from the similar damage data set; when the number of road damage data with the highest recognition confidence is one, the road damage data with the highest recognition confidence is determined as the target road damage data; when the number of road damage data with the highest recognition confidence is multiple, the target road damage data is determined based on the recognition time of multiple road damage data with the highest recognition confidence.

[0100] In an embodiment of the present invention, when the acquisition device is conducting an inspection, it often recognizes road diseases through a road disease recognition model. The road disease recognition model can be constructed and trained by any deep learning algorithm that can recognize road diseases. The deep learning algorithm can be an algorithm of the YOLO (You Only Look Once) series, a Faster R-CNN algorithm, an EfficientDet, etc. After the road disease recognition model recognizes the road disease in the road image, it can output the image coordinates, recognition confidence, disease type and other data of the road disease. Therefore, the road disease data can include the recognition confidence of the road disease. In the similar disease data set, each road disease data has a corresponding recognition confidence. Therefore, the road disease data in the similar disease data set can be sorted in order of recognition confidence from high to low. A sorted similar disease data set is obtained, and then the number of road disease data with the highest recognition confidence is determined based on the sorted similar disease data set. If there is only one road disease data with the highest recognition confidence, the road disease data with the highest recognition confidence can be determined as the target road disease data. If there are multiple road disease data with the highest recognition confidence, the road disease data with the recognition time closest to the current time can be determined from the multiple road disease data with the highest recognition confidence, and determined as the above target disease data.

[0101] Optionally, before the step of obtaining the target disease data set, multiple disease images of the same type, image coordinates of road diseases in the disease images, and acquisition parameters of the target disease images can also be obtained; the size of the target disease image is normalized to obtain the target disease image; based on the image coordinates and the acquisition parameters, the disease location of the road disease in the target disease image is determined; based on the target disease image and the disease location, the target disease data set is constructed.

[0102] In an embodiment of the present invention, the above-mentioned disease image can be acquired by the above-mentioned acquisition device. When the above-mentioned acquisition device recognizes the road image through the above-mentioned road disease recognition model, if it is recognized that there is a road disease in a certain frame of road image, the above-mentioned frame of road image can be determined as a disease image, and the image coordinates of the road disease in the disease image can be output, and the acquisition parameters of the above-mentioned disease image, the above-mentioned image coordinates and the above-mentioned disease image collected can be uploaded to the above-mentioned road inspection platform.

[0103] The above-mentioned normalization processing may include size normalization processing and dedistortion processing. Generally, in order to save storage space, the above-mentioned size normalization processing may be based on the size parameter of the smallest disease image among all disease images as the normalization basis. For example, if the size parameter of the smallest disease image is 1920*1080, the size parameters of other disease images may be normalized to 1920*1080. After normalization, in order to avoid image distortion caused by the lens characteristics of the image acquisition device (or camera) in the acquisition device, the disease image after size normalization may also be dedistorted to obtain the above-mentioned target disease image.

[0104] After normalization, the location of the road damage in the target image can be determined based on the image coordinates and acquisition parameters. A mapping relationship is established based on the target damage image and the corresponding damage location, thereby constructing the above target damage data set.

[0105] Optionally, in the step of determining the damage position of the road damage in the target damage image based on the image coordinates and the acquisition parameters, the three-dimensional coordinates of the road damage can also be determined based on the image coordinates and the internal reference matrix; the dedistorted three-dimensional coordinates of the road damage can be determined based on the three-dimensional coordinates and a preset dedistortion coefficient; and the damage position of the road damage can be determined based on the dedistorted three-dimensional coordinates of the road damage and the acquisition height.

[0106] In an embodiment of the present invention, the acquisition parameters include an acquisition height and an internal reference matrix. The above-mentioned disease image may be acquired by an image acquisition device deployed in the above-mentioned acquisition device, and the above-mentioned image acquisition device may be a binocular camera. Therefore, each pixel point in the above-mentioned disease image may include a depth coordinate, and the above-mentioned internal reference matrix may include a focal length and an optical center. The three-dimensional coordinates of the above-mentioned road disease can be calculated based on the above-mentioned image coordinates and the above-mentioned internal reference matrix. After identifying the road disease, the above-mentioned road disease recognition model can output the recognition frame of the road disease and the coordinates of the recognition frame, that is, the above-mentioned image coordinates. The above-mentioned image coordinates may include the upper left corner coordinates, the lower left corner coordinates, the upper right corner coordinates, and the lower right corner coordinates.

[0107] Specifically, it is assumed that the original resolution of the above disease image is W original *H original The resolution of the disease image after size normalization is W target *H target , the image coordinates include ltx (upper left x-axis coordinate), lty (upper left y-axis coordinate), rbx (lower right x-axis coordinate) and rby (lower right y-axis coordinate). The normalized image coordinates of the road damage in the damage image after size normalization can be calculated by the image coordinates and the original resolution, including the following calculation formula:

[0108]

[0109] Wherein, the above ltx′ represents the upper left x-axis coordinate in the normalized image coordinates, the above lty′ represents the upper left y-axis coordinate in the normalized image coordinates, the above rbx′ represents the lower right x-axis coordinate in the normalized image coordinates, the above rby′ represents the lower right y-axis coordinate in the normalized image coordinates, the above ltx represents the upper left x-axis coordinate in the image coordinates, the above lty represents the upper left y-axis coordinate in the image coordinates, the above rbx represents the lower right x-axis coordinate in the image coordinates, the above rby represents the lower right y-axis coordinate in the image coordinates, and the above W target It is expressed as the lateral resolution of the disease image after size normalization. target It is expressed as the vertical resolution of the disease image after size normalization. original It is expressed as the lateral resolution of the disease image. original Expressed as the vertical resolution of the disease image.

[0110] After obtaining the normalized image coordinates, the pixel coordinates (u, v) of the center point of the road disease can be calculated according to the normalized image coordinates, and the normalized pixel coordinates (x n ,y n ). It should be noted that, in the embodiment of the present invention, the image coordinates may include the coordinates of the upper left corner, the upper right corner, the lower left corner and the lower right corner, and the image coordinates include the pixel coordinates of the four points, and the pixel coordinates may be the coordinates of a specific point, such as the coordinates of the center point in the image.

[0111] Specifically, the normalized pixel coordinates can be calculated using the following formula:

[0112]

[0113] Among them, the above x n Represents the horizontal coordinate of the normalized pixel coordinate. n represents the ordinate of the normalized pixel coordinate, the u represents the abscissa of the pixel coordinate, the v represents the ordinate of the pixel coordinate, and the f x Expressed as the lateral focal length in the above focal length, the above f y Expressed as the longitudinal focal length in the above focal length, the above c x Expressed as the central abscissa of the optical center, the above c y Expressed as the central ordinate of the optical center.

[0114] After the normalized pixel coordinates are obtained, the three-dimensional coordinates may be calculated based on the normalized pixel coordinates and the depth value.

[0115] Specifically, the above three-dimensional coordinates can be calculated according to the following formula:

[0116] X c = Z·x n

[0117] Y c = Z·y n

[0118] Z c =Z

[0119] Among them, the above X c It is expressed as the horizontal coordinate in the above three-dimensional coordinates. c It is expressed as the vertical coordinate in the above three-dimensional coordinates. c It is represented as the depth coordinate in the above three-dimensional coordinates, and the above Z is represented as the above depth value. n Represents the horizontal coordinate of the normalized pixel coordinate. n The vertical coordinate is expressed as the normalized pixel coordinate.

[0120] After obtaining the above pixel coordinates, the pixel coordinates after dedistortion can also be calculated according to the above pixel coordinates and the above preset dedistortion coefficients. The pixel coordinates after dedistortion can be calculated by the following formula:

[0121]

[0122] Wherein, the above u′ represents the horizontal coordinate in the pixel coordinate after the above distortion, the above v′ represents the vertical coordinate in the pixel coordinate after the above distortion, and the above distortion represents the above preset distortion coefficient, wherein distortion = (p1, p2, k1, k2, k3), wherein k1, k2, k3 are radial distortion coefficients, p1, p2 are tangential distortion coefficients, k represents the original internal parameter matrix, K new It is represented as the internal parameter matrix after dedistortion, u is represented as the horizontal coordinate of the pixel coordinate of the center point of the above road disease, and v is represented as the vertical coordinate of the pixel coordinate of the center point of the above road disease.

[0123] After obtaining the pixel coordinates after de-distortion, the image coordinates of the road damage after de-distortion (i.e., the identification frame coordinates or understood as the region boundary coordinates) can be determined according to the pixel coordinates after de-distortion. Specifically, the image coordinates after de-distortion can be calculated by the following formula:

[0124] x min=min(u′)

[0125] x max =max(u′)

[0126] y min =min(v′)

[0127] y max =max(v′)

[0128] Among them, the above x min It is expressed as the minimum horizontal coordinate in the image coordinate after dedistortion. max It is expressed as the maximum horizontal coordinate in the image coordinate after dedistortion. min It is expressed as the minimum ordinate in the dedistorted image coordinates. max The maximum ordinate in the dedistorted image coordinates, the u′ is represented by the abscissa in the dedistorted pixel coordinates, and the v′ is represented by the ordinate in the dedistorted pixel coordinates.

[0129] After the coordinates of the dedistorted image are calculated, the coordinates of the center point of the road damage in the dedistorted road damage image can be calculated based on the coordinates of the dedistorted image.

[0130] Specifically, the coordinates of the center point of the road damage in the road damage image after the above dedistortion processing can be calculated by the following formula:

[0131]

[0132] The center is the coordinate of the center point, and the pixel_coor is the three-dimensional coordinate. c , Y c and Z c , the above x min It is expressed as the minimum horizontal coordinate in the image coordinate after dedistortion. max It is expressed as the maximum horizontal coordinate in the image coordinate after dedistortion. min It is expressed as the minimum ordinate in the dedistorted image coordinates. max The maximum ordinate in the undistorted image coordinates.

[0133] After the coordinates of the center point are calculated, the arc length from the image acquisition device of the acquisition device to the coordinates of the center point can be calculated according to the coordinates of the center point and the acquisition height, which can be calculated by the following formula:

[0134]

[0135] Among them, the above d represents the above arc length, the above x=center[0], y=center[1], z=center[2] respectively represent the positions of the center point on the x, y, and z axes in the camera coordinate system, and the above h represents the above acquisition height.

[0136] Convert the direction angle of the camera (i.e., the image acquisition device of the acquisition device) to radians:

[0137]

[0138] The longitude and latitude of the camera (i.e., the image acquisition device of the acquisition device) are also converted to radians:

[0139]

[0140] where φ camera and λ camera The latitude and longitude of the camera (i.e. the longitude and latitude of the acquisition device when it identifies the road damage)

[0141] Convert the camera coordinates of the disease (i.e., dedistorted 3D coordinates) to geographic longitude and latitude:

[0142]

[0143] where φ damage Expressed as latitude angle (measured in radians), λ damage It is expressed as longitude angle (its unit of measurement is also radians), R is the radius of the earth, and d is the coordinate of the above center point.

[0144] Finally, convert the radians into degrees to get the above disease location (i.e. the latitude and longitude of the center point of the disease):

[0145]

[0146] like Figure 2 As shown, Figure 2 This is a flow chart of another data processing method provided by an embodiment of the present invention:

[0147] The inspection device at the bottom (i.e., the collection device) inspects the road from bottom to top. During the inspection process, road diseases A0, A1, and A2 are identified. Each time a road disease is identified, the disease information (i.e., road disease data) can be uploaded to the inspection system (i.e., the road inspection system). When the inspection system obtains the above-mentioned road diseases A0, A1, and A2 and the corresponding disease information, it constructs a corresponding disease data set as a target disease data set, and executes any one of the data processing methods provided in the embodiments of the present invention to obtain the target disease data, and returns the target disease data to the inspection device. The inspection device can delete other disease data except the target disease data in the similar disease data set.

[0148] like Figure 3 As shown, an embodiment of the present invention further provides a data processing device, including:

[0149] The first acquisition module 301 is used to acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease;

[0150] A first calculation module 302 is used to traverse and calculate in the target disease data set in the order of the recognition time, based on the recognition time and the disease location, to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease;

[0151] A first determination module 303 is used to determine whether to add the first road disease and the second road disease to a similar disease data set based on the disease distance, the identification time interval and the driving distance;

[0152] The first selection module 304 is used to select target road disease data from the similar disease data set after the traversal calculation is completed.

[0153] Optionally, the first calculation module 302 includes:

[0154] A first extraction submodule, configured to extract road defects from the target defect data set one by one as the first road defect in the order of the identification time;

[0155] A first processing submodule is configured to, when each of the first road defects is extracted, take a road defect in the target defect data set whose identification time is later than that of the first road defect as the second road defect;

[0156] The first selection submodule is used to select the second road defects one by one from the target data set in the order of the recognition time, and calculate the defect distance, recognition time interval and driving distance of the acquisition device between the second road defect and the first road defect based on the recognition time and the defect location.

[0157] Optionally, the road damage data includes the driving speed when the acquisition device identifies the road damage, and the first selection submodule includes:

[0158] A first calculation unit, configured to calculate, based on the identification time, an identification time interval between the second road defect and the first road defect;

[0159] A second calculation unit, configured to calculate a defect distance between the second road defect and the first road defect based on the defect location;

[0160] A third calculation unit, configured to calculate an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0161] A fourth calculation unit is used to calculate the driving distance based on the average driving speed and the identification time interval.

[0162] Optionally, the road damage data includes the driving speed when the acquisition device identifies the road damage, and the first determination module 303 includes:

[0163] A first setting submodule is used to set a speed threshold and a distance threshold based on the type of road damage in the target damage data set;

[0164] A first calculation submodule, configured to calculate an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0165] A first determination submodule, configured to determine a time threshold corresponding to the first road damage and the second road damage based on the average driving speed and the speed threshold;

[0166] A second calculation submodule, configured to calculate a remaining distance interval corresponding to the first road defect and the second road defect based on the defect distance and the travel distance;

[0167] The first adding submodule is used to add the first road disease and the second road disease to the similar disease data set when the identification time interval corresponding to the first road disease and the second road disease is less than the time threshold and the corresponding remaining distance interval is less than the distance threshold.

[0168] Optionally, the road damage data includes the recognition confidence of the road damage, and the first selection module 304 includes:

[0169] A second extraction submodule is used to extract road disease data with the highest recognition confidence from the similar disease data set;

[0170] A second determination submodule is used to determine the road damage data with the highest recognition confidence as the target road damage data when the number of the road damage data with the highest recognition confidence is one;

[0171] The third determination submodule is used to determine the target road damage data according to the identification time of the plurality of road damage data with the highest identification confidence when there are a plurality of road damage data with the highest identification confidence.

[0172] Optionally, the data processing device further includes:

[0173] A second acquisition module is used to acquire a plurality of disease images of the same type, image coordinates of road diseases in the disease images, and acquisition parameters of the target disease image;

[0174] A normalization module, used for normalizing the size of the target disease image to obtain a target disease image;

[0175] A second determination module, configured to determine a defect position of a road defect in the target defect image based on the image coordinates and the acquisition parameters;

[0176] A construction module is used to construct the target disease data set based on the target disease image and the disease location.

[0177] Optionally, the acquisition parameters include an acquisition height and an internal parameter matrix, and the second determination module includes:

[0178] A fourth determination submodule, configured to determine the three-dimensional coordinates of the road damage based on the image coordinates and the internal reference matrix;

[0179] A fifth determination submodule, configured to determine the dedistorted three-dimensional coordinates of the road damage based on the three-dimensional coordinates and a preset dedistortion coefficient;

[0180] The sixth determination submodule is used to determine the defect location of the road defect based on the dedistorted three-dimensional coordinates of the road defect and the acquisition height.

[0181] like Figure 4 As shown, an embodiment of the present invention further provides an electronic device, characterized in that it includes a processor, and the processor can execute any one of the above data processing methods.

[0182] Specifically, it includes a processor 401 and a memory 402, and a computer program for executing the data processing method which is stored in the memory 402 and can be run on the processor 401, wherein:

[0183] The processor 401 runs the computer program of the data processing method stored in the memory 402 to perform the following steps:

[0184] Acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease;

[0185] In the order of the recognition time, based on the recognition time and the disease location, the target disease data set is traversed and calculated to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease;

[0186] Determining whether to add the first road disease and the second road disease to a similar disease dataset based on the disease distance, the identification time interval, and the driving distance;

[0187] After the traversal calculation is completed, target road disease data is selected from the similar disease data set.

[0188] Optionally, the processor 401 performs the traversal calculation in the target disease data set in the order of the recognition time and based on the recognition time and the disease location to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device, including:

[0189] Extracting road defects from the target defect data set one by one as the first road defects in the order of the identification time;

[0190] Whenever one of the first road diseases is extracted, a road disease in the target disease data set whose identification time is later than that of the first road disease is taken as the second road disease;

[0191] The second road defects are selected one by one from the target data set in the order of the recognition time, and based on the recognition time and the defect location, the defect distance between the second road defect and the first road defect, the recognition time interval and the travel distance of the acquisition device are calculated.

[0192] Optionally, the road damage data includes the driving speed when the collection device identifies the road damage, and the processor 401 calculates the damage distance between the second road damage and the first road damage, the identification time interval, and the driving distance of the collection device based on the identification time and the damage location, including:

[0193] Based on the recognition time, calculating the recognition time interval between the second road defect and the first road defect;

[0194] Based on the defect location, calculating the defect distance between the second road defect and the first road defect;

[0195] Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0196] The driving distance is calculated based on the average driving speed and the identification time interval.

[0197] Optionally, the road damage data includes the driving speed when the acquisition device identifies the road damage, and the processor 401 determines whether to add the first road damage and the second road damage to the similar damage data set based on the damage distance, the identification time interval and the driving distance, including:

[0198] Based on the type of road damage in the target damage data set, a speed threshold and a distance threshold are set;

[0199] Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect;

[0200] Determining a time threshold corresponding to the first road defect and the second road defect based on the average driving speed and the speed threshold;

[0201] Based on the disease distance and the travel distance, calculating the remaining distance interval corresponding to the first road disease and the second road disease;

[0202] When the identification time interval corresponding to the first road defect and the second road defect is less than the time threshold, and the corresponding remaining distance interval is less than the distance threshold, the first road defect and the second road defect are added to the similar defect data set.

[0203] Optionally, the road damage data includes the recognition confidence of the road damage, and the processor 401 performs the step of selecting target road damage data from the similar damage data set, including:

[0204] Extracting the road disease data with the highest recognition confidence from the similar disease data set;

[0205] When the number of the road damage data with the highest recognition confidence is one, determining the road damage data with the highest recognition confidence as the target road damage data;

[0206] When there are multiple road damage data with the highest recognition confidence, the target road damage data is determined according to the recognition time of the multiple road damage data with the highest recognition confidence.

[0207] Optionally, before acquiring the target disease dataset, the method executed by the processor 401 further includes:

[0208] Acquire multiple disease images of the same type, image coordinates of road diseases in the disease images, and acquisition parameters of the target disease image;

[0209] Normalizing the size of the target disease image to obtain a target disease image;

[0210] Determining the location of the road disease in the target disease image based on the image coordinates and the acquisition parameters;

[0211] Based on the target disease image and the disease location, the target disease dataset is constructed.

[0212] Optionally, the acquisition parameters include an acquisition height and an internal reference matrix, and the processor 401 determines the location of the road disease in the target disease image based on the image coordinates and the acquisition parameters, including:

[0213] Determining the three-dimensional coordinates of the road damage based on the image coordinates and the internal parameter matrix;

[0214] Determining the dedistorted three-dimensional coordinates of the road defect based on the three-dimensional coordinates and a preset dedistortion coefficient;

[0215] The defect location of the road defect is determined based on the dedistorted three-dimensional coordinates of the road defect and the acquisition height.

[0216] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the computer program implements the various processes of the data processing method or the application-side data processing method provided by the embodiment of the present invention, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0217] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the above-mentioned computer program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the above-mentioned computer-readable storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0218] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A data processing method, characterized in that: The method comprises the following steps: Acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease; In the order of the recognition time, based on the recognition time and the disease location, the target disease data set is traversed and calculated to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease; Determining whether to add the first road disease and the second road disease to a similar disease dataset based on the disease distance, the identification time interval, and the driving distance; After the traversal calculation is completed, target road disease data is selected from the similar disease data set.

2. The data processing method according to claim 1, characterized in that: The step of traversing and calculating in the target disease data set in the order of the recognition time and based on the recognition time and the disease location to obtain the disease distance between the first road disease and the second road disease, the recognition time interval, and the travel distance of the collection device includes: Extracting road defects from the target defect data set one by one as the first road defects in the order of the identification time; Whenever one of the first road diseases is extracted, a road disease in the target disease data set whose identification time is later than that of the first road disease is taken as the second road disease; The second road defects are selected one by one from the target data set in the order of the recognition time, and based on the recognition time and the defect location, the defect distance between the second road defect and the first road defect, the recognition time interval and the travel distance of the acquisition device are calculated.

3. The data processing method according to claim 2, characterized in that: The road damage data includes the driving speed when the collection device identifies the road damage, and the calculation of the damage distance between the second road damage and the first road damage, the recognition time interval, and the driving distance of the collection device based on the recognition time and the damage location includes: Based on the recognition time, calculating the recognition time interval between the second road defect and the first road defect; Based on the defect location, calculating the defect distance between the second road defect and the first road defect; Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect; The driving distance is calculated based on the average driving speed and the identification time interval.

4. The data processing method according to claim 1, characterized in that: The road damage data includes the driving speed when the acquisition device identifies the road damage, and determining whether to add the first road damage and the second road damage to the similar damage data set based on the damage distance, the identification time interval and the driving distance includes: Based on the type of road damage in the target damage data set, a speed threshold and a distance threshold are set; Calculating an average driving speed based on the driving speed corresponding to the first road defect and the driving speed corresponding to the second road defect; Determining a time threshold corresponding to the first road defect and the second road defect based on the average driving speed and the speed threshold; Based on the disease distance and the travel distance, calculating the remaining distance interval corresponding to the first road disease and the second road disease; When the identification time interval corresponding to the first road defect and the second road defect is less than the time threshold, and the corresponding remaining distance interval is less than the distance threshold, the first road defect and the second road defect are added to the similar defect data set.

5. The data processing method according to claim 1, characterized in that: The road damage data includes the recognition confidence of the road damage, and the target road damage data is selected from the similar damage data set, including: Extracting the road disease data with the highest recognition confidence from the similar disease data set; When the number of the road damage data with the highest recognition confidence is one, determining the road damage data with the highest recognition confidence as the target road damage data; When there are multiple road damage data with the highest recognition confidence, the target road damage data is determined according to the recognition time of the multiple road damage data with the highest recognition confidence.

6. The data processing method according to any one of claims 1 to 5, characterized in that: Before obtaining the target disease dataset, the method further includes: Acquire multiple disease images of the same type, image coordinates of road diseases in the disease images, and acquisition parameters of the target disease image; Normalizing the size of the target disease image to obtain a target disease image; Determining the location of the road disease in the target disease image based on the image coordinates and the acquisition parameters; Based on the target disease image and the disease location, the target disease dataset is constructed.

7. The data processing method according to claim 6, characterized in that: The acquisition parameters include an acquisition height and an internal reference matrix, and determining the defect position of the road defect in the target defect image based on the image coordinates and the acquisition parameters includes: Determining the three-dimensional coordinates of the road damage based on the image coordinates and the internal parameter matrix; Determining the dedistorted three-dimensional coordinates of the road defect based on the three-dimensional coordinates and a preset dedistortion coefficient; The defect location of the road defect is determined based on the dedistorted three-dimensional coordinates of the road defect and the acquisition height.

8. A data processing device, characterized in that: The data processing device comprises: A first acquisition module is used to acquire a target disease data set, wherein the target disease data set includes a plurality of road disease data of the same type, wherein the road disease data includes the recognition time of the road disease and the disease location of the road disease, wherein the recognition time is the time when the acquisition device recognizes the road disease; A first calculation module is used to traverse and calculate in the target disease data set in the order of the recognition time, based on the recognition time and the disease location, to obtain the disease distance between the first road disease and the second road disease, the recognition time interval and the travel distance of the collection device, wherein the recognition time of the first road disease is earlier than the recognition time of the second road disease; A first determination module, configured to determine whether to add the first road disease and the second road disease to a similar disease data set based on the disease distance, the identification time interval, and the driving distance; The first selection module is used to select target road disease data from the similar disease data set after the traversal calculation is completed.

9. An electronic device, characterized in that: include: 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 data processing method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the data processing method according to any one of claims 1 to 7 are implemented.

Images