A license plate correction method, device, terminal and computer-readable storage medium

Abstract

The present invention provides a license plate correction method, device, terminal and computer-readable storage medium. In the license plate correction method, the running track of a target vehicle in a preset area and a preset time period is obtained; the target vehicle has corresponding license plate detection information and vehicle detection information; based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate; in response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected. In the present application, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate through the running track, license plate detection information and vehicle detection information of the target vehicle, and then the license plate detection information corresponding to the abnormal license plate is corrected to improve the accuracy of the vehicle license plate information.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a license plate correction method, device, terminal and computer-readable storage medium. Background Art

[0002] As we all know, the license plate number of a vehicle is the identification of the vehicle. License plate recognition technology can realize automatic registration and verification of the vehicle without making any changes to the vehicle. This technology has been applied to various occasions such as highway toll collection, parking management, weighing system, traffic induction, traffic law enforcement, highway inspection, vehicle dispatching, and vehicle detection. In the current construction of smart cities, license plate recognition has become a top priority. Traditional license plate recognition is to identify vehicle license plate number and color information through technologies such as license plate extraction, image preprocessing, feature extraction, and license plate character recognition. However, the recognition rate of license plate recognition has not yet been 100% perfect. The reason is that its recognition rate is affected by many factors: such as the shooting angle of the image acquisition device, driving speed, external ambient light, occlusion, license plate contamination, etc., which causes the license plate recognition error. Obviously, there is still a lot of room for improvement in the current license plate recognition system. Summary of the invention

[0003] The main technical problem solved by the present invention is to provide a license plate correction method, device, terminal and computer-readable storage medium to solve the problem of low license plate correction accuracy in the prior art.

[0004] In order to solve the above technical problems, the first technical solution adopted by the present invention is: to provide a license plate correction method, the license plate correction method comprising:

[0005] Obtain the running track of the target vehicle in a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information;

[0006] Determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle;

[0007] In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected.

[0008] The operation track of the target vehicle in the preset area and the preset time period is obtained, including:

[0009] Obtain the running tracks of all detected vehicles corresponding to the preset area and preset time period, each detected vehicle has corresponding license plate detection information and vehicle detection information;

[0010] Each of all detected vehicles is taken as a target vehicle in turn.

[0011] Wherein, based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0012] The abnormal license plate detection model is used to determine the license plate detection result of the target vehicle's license plate detection information based on the running trajectory, license plate detection information and vehicle detection information corresponding to the target vehicle; the license plate detection result is an abnormal license plate or a normal license plate;

[0013] Among them, the abnormal license plate detection model includes multiple detection modules.

[0014] The abnormal license plate detection model is used to determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running trajectory, license plate detection information and vehicle detection information corresponding to the target vehicle, including:

[0015] Each detection module performs abnormal license plate detection based on the running track, license plate detection information and vehicle detection information corresponding to the target vehicle, and obtains the initial detection results corresponding to each detection module;

[0016] The initial detection results corresponding to each detection module are weighted and summed to obtain the license plate detection result corresponding to the license plate detection information of the target vehicle.

[0017] Wherein, the license plate detection information includes the detected license plate number;

[0018] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0019] Compare the detected license plate number of the target vehicle with the string regularization rules of each preset license plate number in the database;

[0020] In response to the detected license plate number not matching any of the character string regularization rules of each preset license plate number in the database, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0021] Among them, the license plate detection information includes the detected license plate number and the detected license plate type;

[0022] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0023] Determine whether the detection license plate number corresponding to the target vehicle matches the detection license plate type;

[0024] In response to the detection license plate number corresponding to the target vehicle not matching the detection license plate type, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0025] Among them, the running trajectory includes the trajectory points that the target vehicle has passed through successively;

[0026] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0027] Determine the detection running speed and / or detection time difference of the target vehicle based on the spatiotemporal information of each track point in the running track of the target vehicle;

[0028] In response to the detection running speed and / or detection time difference corresponding to the target vehicle not meeting the corresponding preset requirements, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0029] Wherein, the license plate detection information includes the detected license plate number;

[0030] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0031] Recognize the license plate number of the target vehicle by the first device and the second device respectively, and obtain two detection license plate numbers corresponding to the target vehicle;

[0032] In response to the inconsistency between the detected license plate numbers of the target vehicle respectively detected by the first device and the second device, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0033] Wherein, the license plate detection information includes the detected license plate number and / or the detected license plate type;

[0034] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0035] Determine whether the target vehicle has a corresponding detection license plate number and detection license plate type;

[0036] In response to the target vehicle not having a corresponding detected license plate number or detected license plate type, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0037] Among them, the running trajectory includes the trajectory points that the target vehicle has passed through successively;

[0038] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0039] Based on the running track of the target vehicle, the number of times the target vehicle appears at each track point within a preset sub-period is obtained by counting; the preset sub-period is included in the preset period;

[0040] In response to the number of times the target vehicle appears at the trajectory point within the preset sub-period not meeting the preset number, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0041] Wherein, based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0042] Compare the license plate detection information of the target vehicle with each preset license plate information in the database;

[0043] In response to the license plate detection information of the target vehicle being different from each preset license plate information in the database, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0044] Wherein, based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0045] In response to the license plate detection information of the target vehicle being identical to a preset license plate information in the database, the vehicle detection information of the target vehicle is compared with the preset vehicle information corresponding to the preset license plate information;

[0046] In response to the vehicle detection information of the target vehicle being different from the preset vehicle information corresponding to the preset license plate information, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0047] Wherein, the target vehicle has a corresponding vehicle-related object;

[0048] Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0049] Based on the vehicle-associated object corresponding to the target vehicle, the license plate detection information, and the vehicle detection information, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate.

[0050] Wherein, based on the vehicle-related object corresponding to the target vehicle, the license plate detection information and the vehicle detection information, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0051] Determine whether the vehicle-associated object corresponding to the target vehicle is consistent with the vehicle-associated objects corresponding to other detected vehicles other than the target vehicle corresponding to the preset area and the preset time period;

[0052] In response to the vehicle-associated object of the target vehicle being consistent with the vehicle-associated object of at least one detection vehicle, determining the vehicle similarity between the target vehicle and the detection vehicle corresponding to the same vehicle-associated object based on the license plate detection information and the vehicle detection information corresponding to the target vehicle and the detection vehicle respectively;

[0053] In response to the vehicle similarity being greater than the first threshold, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0054] Wherein, the license plate detection information includes the detected license plate number;

[0055] Based on the vehicle-related object, the license plate detection information, and the vehicle detection information corresponding to the target vehicle, determining whether the license plate detection information of the target vehicle is an abnormal license plate includes:

[0056] Based on the vehicle-related object of the target vehicle, searching the database for a preset license plate number corresponding to the vehicle-related object;

[0057] In response to the detection license plate number corresponding to the vehicle-associated object of the target vehicle being inconsistent with the preset license plate number, determining the vehicle similarity corresponding to the detection license plate number and the preset license plate number based on the license plate detection information corresponding to the target vehicle, the vehicle detection information, and the vehicle registration information corresponding to the preset license plate number;

[0058] In response to the vehicle similarity being greater than the second threshold, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0059] In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected, including:

[0060] In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected using a license plate correction model to obtain a license plate correction result of the target vehicle;

[0061] Among them, the license plate correction model includes multiple license plate correction modules.

[0062] The license plate correction model is used to correct the license plate detection information of the target vehicle to obtain the license plate correction result of the target vehicle, including:

[0063] Based on each license plate correction module, the license plate detection information of the target vehicle is corrected respectively to obtain the license plate correction information output by each license plate correction module;

[0064] The license plate correction information output by each license plate correction module is weighted and summed to obtain the license plate correction result of the target vehicle.

[0065] Among them, the license plate correction method includes:

[0066] Filter out the detected vehicles corresponding to abnormal license plates from all detected vehicles corresponding to the preset area and preset time period to obtain a normal license plate data set corresponding to the preset area and preset time period; the normal license plate data set includes the detected vehicles corresponding to the normal license plates.

[0067] Among them, the license plate detection information includes the detected license plate number and the detected license plate type.

[0068] Correct the license plate detection information of the target vehicle, including:

[0069] Based on the detected license plate numbers and detected license plate types corresponding to each detected vehicle in the normal license plate data set, a first relationship table corresponding to the license plate numbers and the license plate types is obtained;

[0070] Compare the detected license plate number of the target vehicle with each license plate number in the first relationship table;

[0071] The detected license plate type of the target vehicle is corrected based on the license plate type corresponding to the license plate number of the same type as the detected license plate number.

[0072] Among them, the license plate detection information includes the detected license plate number and the detected license plate type, and the target vehicle and the detected vehicle both have corresponding vehicle association objects;

[0073] Correct the license plate detection information of the target vehicle, including:

[0074] Based on the license plate detection information corresponding to each detected vehicle in the normal license plate data set and the vehicle associated object, a second relationship table corresponding to the license plate information and the associated object is obtained;

[0075] Compare the vehicle associated object of the target vehicle with each associated object in the second relationship table;

[0076] Based on the license plate information corresponding to the associated object that is the same as the vehicle associated object of the target vehicle, the detected license plate number and / or the detected license plate type corresponding to the target vehicle is corrected.

[0077] Among them, the license plate detection information includes the detected license plate number and the detected license plate type.

[0078] Correct the license plate detection information of the target vehicle, including:

[0079] Compare the detected license plate number of the target vehicle with the preset license plate number in the third relationship table; the third relationship table is a pre-constructed correspondence table between the preset license plate number and the preset license plate type;

[0080] The detected license plate type of the target vehicle is corrected based on a preset license plate type corresponding to a preset license plate number of the same type as the detected license plate number of the target vehicle.

[0081] Among them, the license plate detection information includes the detected license plate number and the detected license plate type.

[0082] Correct the license plate detection information of the target vehicle, including:

[0083] Determine the vehicle similarity between the target vehicle and the registered vehicles based on the license plate detection information and vehicle detection information corresponding to the target vehicle and the license plate information and vehicle information corresponding to each registered vehicle in the database;

[0084] In response to the vehicle similarity reaching a third threshold, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected based on the license plate information of the registered vehicle corresponding to the vehicle similarity.

[0085] Among them, the license plate detection information includes the detected license plate number and the detected license plate type.

[0086] Correct the license plate detection information of the target vehicle, including:

[0087] Determine the vehicle similarity between the detected vehicle and the target vehicle based on the license plate detection information and vehicle detection information corresponding to the detected vehicle and the target vehicle in the neighboring space-time corresponding to the preset area and the preset time period;

[0088] In response to the vehicle similarity reaching a fourth threshold, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity.

[0089] Among them, the license plate detection information includes the detected license plate number and the detected license plate type.

[0090] Correct the license plate detection information of the target vehicle, including:

[0091] Based on the running track of the target vehicle, search in the preset map to determine the track points that the target vehicle must pass through;

[0092] Based on the license plate detection information and vehicle detection information corresponding to each detection vehicle and the target vehicle collected at the necessary trajectory points, the vehicle similarity between the detection vehicle and the target vehicle is determined;

[0093] In response to the vehicle similarity reaching a fifth threshold, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity.

[0094] In order to solve the above technical problems, the third technical solution adopted by the present invention is: to provide a license plate correction device, the license plate correction device comprising:

[0095] An acquisition module is used to acquire the running track of a target vehicle within a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information;

[0096] An analysis module, used to determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle;

[0097] The correction module is used to correct the license plate detection information of the target vehicle in response to the license plate detection information of the target vehicle being an abnormal license plate.

[0098] To solve the above technical problems, the third technical solution adopted by the present invention is: to provide a terminal, the terminal includes a memory, a processor and a computer program stored in the memory and running on the processor, the processor is used to execute program data to implement the steps in the above license plate correction method.

[0099] To solve the above technical problems, the fourth technical solution adopted by the present invention is: providing a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the above license plate correction method are implemented.

[0100] The beneficial effects of the present invention are as follows: different from the prior art, a license plate correction method, device, terminal and computer-readable storage medium are provided, in which the running track of the target vehicle in a preset area and a preset time period is obtained; the target vehicle has corresponding license plate detection information and vehicle detection information; based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate; in response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected. In the present application, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate through the running track, license plate detection information and vehicle detection information of the target vehicle, and then the license plate detection information corresponding to the abnormal license plate is corrected, thereby improving the accuracy of the vehicle license plate information. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments 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 creative work.

[0102] Figure 1 It is a flow chart of the license plate correction method provided by the present invention;

[0103] Figure 2It is a schematic framework diagram of an embodiment of the license plate correction device provided by the present invention;

[0104] Figure 3 It is a schematic framework diagram of an embodiment of the terminal provided by the present invention;

[0105] Figure 4 It is a schematic framework diagram of an embodiment of the computer-readable storage medium provided by the present invention. Detailed implementation manners

[0106] The following will combine the accompanying drawings of the specification to elaborate in detail on the solutions of the embodiments of the present application.

[0107] In the following description, specific details such as specific system structures, interfaces, and technologies are proposed for the purpose of illustration rather than limitation, so as to thoroughly understand the present application.

[0108] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after. In addition, "multiple" in this article means two or more than two.

[0109] To enable those skilled in the art to better understand the technical solutions of the present invention, the following will further elaborate in detail on a license plate correction method provided by the present invention in combination with the drawings and specific implementation manners.

[0110] Unless otherwise defined, all technical and scientific terms used in this article have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used in this article are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0111] Before further elaborating on the embodiments of the present application, the nouns and terms involved in the embodiments of the present application will be described. The nouns and terms involved in the embodiments of the present application are applicable to the following explanations.

[0112] The similarity of license plate numbers is the similarity between two recognized license plate numbers. Specifically, it includes but is not limited to analyzing the longest common subsequence between two license plate numbers. For example, the longest common subsequence between Zhe A1**45D and Zhe A1*45D is Zhe A1*45D, with a length of 7 digits, so the similarity of the license plate numbers is 7 / 8.

[0113] The near neighbor space-time is to divide a preset area using the Geohash algorithm, and collect 9 sub-blocks corresponding to the longitude and latitude of the data collection device of the target as the near neighbor space, and collect the preset time periods before and after the time of the target as the near neighbor time. The near neighbor space and the near neighbor time correspond to the near neighbor space-time.

[0114] Among them, Geohash is a method of spatial indexing. Its basic principle is to understand the earth as a two-dimensional plane. By encoding the two-dimensional spatial longitude and latitude data into a string, the plane can be recursively decomposed into smaller sub-blocks. Each sub-block has the same code within a certain longitude and latitude range.

[0115] Box plot, also known as box-and-whisker plot, box-style plot or box-and-whisker plot, is a statistical chart used to show the dispersion of a set of data. It is named because of its box-like shape. It is also often used in various fields, and is common in quality management. It is mainly used to reflect the characteristics of the original data distribution, and can also compare the distribution characteristics of multiple groups of data. The method of drawing a box-and-whisker plot is: first find the maximum value, minimum value, median and two quartiles of a set of data; then, connect the two quartiles to draw a box; then connect the maximum value and minimum value to the box, and the median is in the middle of the box.

[0116] The license plate correction method provided in the embodiment of the present application can be implemented by the server or the terminal alone, or by the server and the terminal in collaboration. In some embodiments, the terminal or the server can implement the license plate correction method provided in the embodiment of the present application by running a computer program. For example, the computer program can be a native program or software module in the operating system; it can be a native application (APP, Application), that is, a program that needs to be installed in the operating system to run, such as a client that supports virtual scenes, such as a game APP; it can also be a small program, that is, a program that only needs to be downloaded to a browser environment to run; it can also be a small program that can be embedded in any APP. In short, the above-mentioned computer program can be an application, module or plug-in in any form.

[0117] The license plate correction method provided in the embodiment of the present application is explained below using server implementation as an example.

[0118] See also Figure 1 , Figure 1 It is a flow chart of the license plate correction method provided by the present invention.

[0119] The present embodiment provides a license plate correction method, which includes the following steps.

[0120] S1: Obtain the running track of the target vehicle in a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information.

[0121] S2: Based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determine whether the license plate detection information of the target vehicle is an abnormal license plate.

[0122] S3: In response to the license plate detection information of the target vehicle being an abnormal license plate, correcting the license plate detection information of the target vehicle.

[0123] Specifically, in step S1, the running track of the target vehicle in a preset area and a preset time period is obtained; the specific implementation method of the target vehicle having corresponding license plate detection information and vehicle detection information is as follows.

[0124] The running tracks corresponding to all the detection vehicles corresponding to the preset area and the preset time period are obtained, and each detection vehicle has corresponding license plate detection information and vehicle detection information. The license plate detection information includes the detection license plate number and / or the detection license plate type; the vehicle detection information includes at least one of the vehicle type, body color, vehicle brand, vehicle annual inspection mark number, and vehicle front window pendant. The running track includes the track points that the target vehicle has passed through successively.

[0125] In one embodiment, data information is collected by various data collection devices arranged in a preset area. The data collection device may be an image collection device, an ETC collection device, or an IMSI collection device. In this embodiment, the data collection device is an image collection device as an example, and the data information is a captured image. The full name of ETC in English is Electronic Toll Collection, and the full name in Chinese is Non-Stop Toll Collection System. The full name of IMSI in English is International Mobile Subscriber Identification Number, and the full name in Chinese is International Mobile Target Identification Number.

[0126] In a specific embodiment, each image acquisition device arranged in a preset area acquires the captured images containing the vehicle within a preset time period. By performing target detection and clustering on all captured images acquired within the preset area and the preset time period, the image subsets corresponding to each detected vehicle, the detected license plate number, and the detected license plate type are obtained. The captured images are sorted based on the acquisition time of the captured images containing the detected vehicle in the image subset of the detected vehicle, and the running track of the detected vehicle is generated based on the captured images obtained by sorting and the latitude and longitude information of the image acquisition device corresponding to the captured images.

[0127] Each of all detected vehicles is taken as a target vehicle in turn.

[0128] Specifically, in step S2, based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, a specific implementation method for determining whether the license plate detection information of the target vehicle is an abnormal license plate is as follows.

[0129] In one embodiment, an abnormal license plate detection model is used to determine a license plate detection result of the target vehicle's license plate detection information based on the target vehicle's corresponding running trajectory, license plate detection information, and vehicle detection information; the license plate detection result is an abnormal license plate or a normal license plate.

[0130] In a specific embodiment, the abnormal license plate detection model includes multiple detection modules. Each detection module performs abnormal license plate detection based on the running track, license plate detection information and vehicle detection information corresponding to the target vehicle, and obtains the initial detection results corresponding to each detection module; the initial detection results corresponding to each detection module are weighted summed to obtain the license plate detection result corresponding to the license plate detection information of the target vehicle.

[0131] In one embodiment, the detected license plate number of the target vehicle is compared with the string regularization rules of each preset license plate number in the database. For example, the string regularization rule of the preset license plate number corresponding to non-energy vehicles issued by various provinces and cities is: province + region code + five-digit code; the string regularization rule of the preset license plate number corresponding to new energy small vehicles issued by various provinces and cities is: province + region code + whether it is a pure electric sign code + five-digit code, etc.

[0132] In response to the detected license plate number not matching any of the string regularization rules of each preset license plate number in the database, it is determined that the license plate detection information of the target vehicle is an abnormal license plate. The license plate detection information of the target vehicle is an abnormal license plate of the first category.

[0133] In one embodiment, the first detection module can be used to compare the detected license plate number of the target vehicle with the string regularization rules of each preset license plate number in the database, and the first detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the first detection module is α1. The weight α1 corresponding to the first detection module is obtained by pre-training the abnormal license plate detection model.

[0134] In one embodiment, it is determined whether the detection license plate number corresponding to the target vehicle matches the detection license plate type. In response to the detection license plate number corresponding to the target vehicle not matching the detection license plate type, it is determined that the license plate detection information of the target vehicle is an abnormal license plate. The license plate detection information of the target vehicle is an abnormal license plate of the second category.

[0135] Among them, the detected license plate types can be pure electric new energy small cars, pure electric new energy large cars, hybrid new energy small cars, trailers, police cars, training cars, etc.

[0136] For example, among the currently issued vehicle license plates, the license plate number of a pure electric new energy small car license plate has eight digits, and the third digit is "D, A, B, C"; the license plate number of a pure electric new energy large car license plate has eight digits, and the eighth digit is "D"; the license plate number of a hybrid new energy small car license plate has eight digits, and the third digit is "F, G"; the last digit of a trailer is "gua"; the last digit of a police car is "jing"; the last digit of a training car is "xue", etc.

[0137] In a specific embodiment, when the detected license plate number of the target vehicle is "Jing*****jing", the detected license plate type of the target vehicle is "training car", while the license plate type of the target vehicle should be "police car", then the detected license plate number of the target vehicle does not match the detected license plate type, and it is determined that the license plate detection information of the target vehicle is an abnormal license plate of the second category.

[0138] In an embodiment, a second detection module can be used to determine whether the detected license plate number of the target vehicle matches the detected license plate type. The second detection module outputs an initial detection result of the license plate detection information of the target vehicle and the confidence level of this initial detection result. Among them, the weight corresponding to the second detection module is α2. The weight α2 corresponding to the second detection module is obtained when the abnormal license plate detection model is pre-trained.

[0139] In an embodiment, based on the spatio-temporal information of each trajectory point in the running trajectory of the target vehicle, the detected running speed and / or the detected time difference of the target vehicle are determined; in response to the detected running speed and / or the detected time difference of the target vehicle not meeting the corresponding preset requirements, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0140] In a specific embodiment, each trajectory point in the running trajectory of the target vehicle has a corresponding acquisition time and acquisition location. The detected running speed of the target vehicle is determined based on the shortest driving distance between the acquisition locations of adjacent trajectory points of the target vehicle in the preset map or the straight-line distance between the acquisition locations of the trajectory points and the time difference between the acquisition times. Among them, the acquisition location is the longitude and latitude information of the image acquisition device. When the detected running speed of the target vehicle is greater than the speed threshold, it is determined that the license plate detection information of the target vehicle is an abnormal license plate of the third category.

[0141] In the running trajectory of the target vehicle, find the n detection time differences corresponding to the target vehicle passing through adjacent trajectory points A and trajectory points B. Compare the n detection time differences with the time difference thresholds respectively. In response to the detection time difference being greater than the time difference threshold, it is determined that the running trajectory of the target vehicle is a discontinuous trajectory, and the detection time difference is eliminated. A box plot and other methods are used to calculate the abnormal lower limit abnormal time difference based on the remaining detection time differences in the n detection time differences. Each remaining detection time difference in the n detection time differences is compared with the abnormal lower limit abnormal time difference. In response to the detection time differences being less than the abnormal lower limit abnormal time difference, it is determined that the license plate detection information of the target vehicle is an abnormal license plate of the third category.

[0142] In one embodiment, a third detection module can be used based on the spatiotemporal information of each track point in the running track of the target vehicle, and the third detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the third detection module is α3. The weight α3 corresponding to the third detection module is obtained by pre-training the abnormal license plate detection model.

[0143] In one embodiment, the license plate number of the target vehicle is identified by the first device and the second device respectively, and two detection license plate numbers corresponding to the target vehicle are obtained; in response to the detection license plate numbers of the target vehicle detected by the first device and the second device respectively being inconsistent, the license plate detection information of the target vehicle is determined to be an abnormal license plate of the fourth category. The first device may be a front-end image acquisition device, and the second device may be a back-end server that receives and analyzes data.

[0144] In one embodiment, the fourth detection module can be used to identify and compare the license plate number of the target vehicle through the first device and the second device respectively, and the fourth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the fourth detection module is α4. The weight α4 corresponding to the fourth detection module is obtained by pre-training the abnormal license plate detection model.

[0145] In one embodiment, it is determined whether the target vehicle has a corresponding detection license plate number and detection license plate type. In response to the target vehicle not having a corresponding detection license plate number or detection license plate type, the license plate detection information of the target vehicle is determined to be an abnormal license plate.

[0146] In one embodiment, the fifth detection module can be used to determine whether the target vehicle has the corresponding detection license plate number and detection license plate type, and the fifth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the fifth detection module is α5. The weight α5 corresponding to the fifth detection module is obtained by pre-training the abnormal license plate detection model.

[0147] In one embodiment, based on the running track of the target vehicle, the number of times the target vehicle appears at each track point within a preset sub-period is statistically obtained; the preset sub-period is included in the preset period. In response to the number of times the target vehicle appears at the track point within the preset sub-period not meeting the preset number, the license plate detection information of the target vehicle is determined to be an abnormal license plate.

[0148] In one embodiment, the sixth detection module can be used to obtain the number of times the target vehicle appears at each track point within a preset sub-period based on the running track of the target vehicle, and the sixth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the sixth detection module is α6. The weight α6 corresponding to the sixth detection module is obtained by pre-training the abnormal license plate detection model.

[0149] In one embodiment, the license plate detection information of the target vehicle is compared with each preset license plate information in the database. In response to the license plate detection information of the target vehicle being different from each preset license plate information in the database, the license plate detection information of the target vehicle is determined to be an abnormal license plate. The database contains a plurality of registered license plate information. The registered license plate information is used as the preset license plate information. The registered license plate information includes a registered license plate number and a registered license plate type.

[0150] In one embodiment, in response to the detected license plate number of the target vehicle being different from all registered license plate numbers in the database, the detected license plate number of the target vehicle is determined to be an abnormal license plate.

[0151] In one embodiment, the seventh detection module can be used to compare the license plate detection information of the target vehicle with each preset license plate information in the database, and the seventh detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the seventh detection module is α7. The weight α7 corresponding to the seventh detection module is obtained by pre-training the abnormal license plate detection model.

[0152] In one embodiment, in response to the target vehicle's license plate detection information being identical to a preset license plate information in a database, the target vehicle's vehicle detection information is compared with the preset vehicle information corresponding to the preset license plate information. In response to the target vehicle's vehicle detection information being different from the preset vehicle information corresponding to the preset license plate information, the target vehicle's license plate detection information is determined to be an abnormal license plate.

[0153] In one embodiment, when the license plate detection information of the target vehicle is the same as a preset license plate information in the database, the eighth detection module can be used to compare the vehicle detection information of the target vehicle with the preset vehicle information corresponding to the preset license plate information, and the eighth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the eighth detection module is α8. The weight α8 corresponding to the eighth detection module is obtained by pre-training the abnormal license plate detection model.

[0154] In one embodiment, the target vehicle has a corresponding vehicle-related object. Based on the vehicle-related object corresponding to the target vehicle, the license plate detection information and the vehicle detection information, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate.

[0155] In a specific embodiment, it is determined whether the vehicle-associated object corresponding to the target vehicle is consistent with the vehicle-associated objects corresponding to other detected vehicles other than the target vehicle corresponding to the preset area and preset time period. In response to the vehicle-associated object of the target vehicle being consistent with the vehicle-associated object of at least one detected vehicle, the vehicle similarity between the target vehicle and the detected vehicle corresponding to the same vehicle-associated object is determined based on the license plate detection information and vehicle detection information corresponding to the target vehicle and the detected vehicle, respectively. In response to the vehicle similarity being greater than a first threshold, the license plate detection information of the target vehicle is determined to be an abnormal license plate.

[0156] In one embodiment, the ninth detection module can be used to determine whether the vehicle-related object corresponding to the target vehicle is consistent with the vehicle-related objects corresponding to other detection vehicles other than the target vehicle corresponding to the preset area and preset time period, and determine the vehicle similarity between the target vehicle and the detection vehicle corresponding to the same vehicle-related object based on the license plate detection information and vehicle detection information corresponding to the target vehicle and the detection vehicle respectively, and the ninth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. Among them, the weight corresponding to the ninth detection module is α9. The weight α9 corresponding to the ninth detection module is obtained by pre-training the abnormal license plate detection model.

[0157] In a specific embodiment, based on the vehicle-related object of the target vehicle, a preset license plate number corresponding to the vehicle-related object is searched in a database.

[0158] In response to the inconsistency between the detected license plate number corresponding to the vehicle-associated object of the target vehicle and the preset license plate number, the vehicle similarity between the detected license plate number and the preset license plate number is determined based on the license plate detection information corresponding to the target vehicle, the vehicle detection information and the vehicle registration information corresponding to the preset license plate number.

[0159] In response to the vehicle similarity being greater than the second threshold, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

[0160] In one embodiment, the tenth detection module can be used to search the database for a preset license plate number corresponding to the vehicle-associated object based on the vehicle-associated object of the target vehicle, and determine the vehicle similarity between the detected license plate number and the preset license plate number based on the license plate detection information corresponding to the target vehicle, the vehicle detection information, and the vehicle registration information corresponding to the preset license plate number. The tenth detection module outputs the initial detection result of the license plate detection information of the target vehicle and the confidence of the initial detection result. The weight corresponding to the tenth detection module is α 10 The weight α corresponding to the tenth detection module 10 It is obtained by pre-training the abnormal license plate detection model.

[0161] In one embodiment, the running track, vehicle detection information, and license plate detection information of the target vehicle are obtained, and the running track, vehicle detection information, and license plate detection information of the target vehicle are respectively detected by the first detection module, the second detection module, the third detection module, the fourth detection module, the fifth detection module, the sixth detection module, the seventh detection module, the eighth detection module, the ninth detection module, and the tenth detection module, to obtain the initial detection result corresponding to each detection module and the confidence of the initial detection result. The weights, initial detection results, and detection results of the initial detection result confidence of each detection module are fused to obtain the license plate detection result corresponding to the license plate detection information of the target vehicle. Among them, the license plate detection result is an abnormal license plate or a normal license plate.

[0162] In one embodiment, the detected vehicles corresponding to abnormal license plates in all detected vehicles corresponding to the preset area and preset time period are filtered out to obtain a normal license plate data set corresponding to the preset area and preset time period; the normal license plate data set includes the detected vehicles corresponding to the normal license plates.

[0163] Specifically, in response to the license plate detection information of the target vehicle being an abnormal license plate in step S3, a specific implementation method of correcting the license plate detection information of the target vehicle is as follows.

[0164] In one embodiment, in response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected using a license plate correction model to obtain a license plate correction result of the target vehicle.

[0165] In a specific embodiment, the license plate correction model includes a plurality of license plate correction modules. Based on each license plate correction module, the license plate detection information of the target vehicle is corrected to obtain the license plate correction information output by each license plate correction module; the license plate correction information output by each license plate correction module is weighted and summed to obtain the license plate correction result of the target vehicle.

[0166] In one embodiment, based on the detected license plate numbers and detected license plate types corresponding to each detected vehicle in the normal license plate data set, a first relationship table corresponding to the license plate numbers and license plate types is obtained; the detected license plate number of the target vehicle is compared with each license plate number in the first relationship table; based on the license plate type corresponding to the license plate number of the same type as the detected license plate number, the detected license plate type of the target vehicle is corrected.

[0167] In one embodiment, the first license plate correction module can be used to correct the abnormal license plate based on the detected license plate number and the detected license plate type corresponding to each detected vehicle in the normal license plate data set, and the first license plate correction module outputs the license plate correction information corresponding to the target vehicle and the license plate correction information confidence. Among them, the weight corresponding to the first license plate correction module is β1. The weight β1 corresponding to the first license plate correction module is obtained by pre-training the license plate correction model.

[0168] In one embodiment, both the target vehicle and the detected vehicle have corresponding vehicle-related objects. Based on the license plate detection information and vehicle-related objects corresponding to each detected vehicle in the normal license plate data set, a second relationship table corresponding to the license plate information and the related objects is obtained; the vehicle-related object of the target vehicle is compared with each related object in the second relationship table; based on the license plate information corresponding to the related object that is the same as the vehicle-related object of the target vehicle, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected.

[0169] In one embodiment, a second license plate correction module can be used to correct abnormal license plates based on the license plate detection information corresponding to each detected vehicle in the normal license plate data set and the vehicle-related object, and the second license plate correction module outputs the license plate correction information corresponding to the target vehicle and the license plate correction information confidence. Among them, the weight corresponding to the second license plate correction module is β2. The weight β2 corresponding to the second license plate correction module is obtained by pre-training the license plate correction model.

[0170] In one embodiment, the detected license plate number of the target vehicle is compared with a preset license plate number in a third relationship table; the third relationship table is a pre-constructed correspondence table between preset license plate numbers and preset license plate types. Based on the preset license plate type corresponding to the preset license plate number of the same type as the detected license plate number of the target vehicle, the detected license plate type of the target vehicle is corrected.

[0171] Among them, the preset license plate types include but are not limited to pure electric new energy small cars, pure electric new energy large cars, hybrid new energy small cars, hybrid new energy large cars, military vehicles, police cars, etc.

[0172] In one embodiment, a third license plate correction module can be used to compare the detected license plate number of the target vehicle with the preset license plate number in the third relationship table to correct the abnormal license plate, and the third license plate correction module outputs the license plate correction information corresponding to the target vehicle and the confidence of the license plate correction information. Among them, the weight corresponding to the third license plate correction module is β3. The weight β3 corresponding to the third license plate correction module is obtained by pre-training the license plate correction model.

[0173] In one embodiment, based on the license plate detection information corresponding to the target vehicle and the vehicle detection information and the license plate information and vehicle information corresponding to each registered vehicle in the database, the vehicle similarity between the target vehicle and the registered vehicle is determined. In response to the vehicle similarity reaching a third threshold, the detection license plate number and / or the detection license plate type corresponding to the target vehicle is corrected based on the license plate information of the registered vehicle corresponding to the vehicle similarity.

[0174] In one embodiment, the fourth license plate correction module can be used to correct the abnormal license plate based on the license plate detection information corresponding to the target vehicle and the vehicle detection information and the license plate information and vehicle information corresponding to each registered vehicle in the database, and the fourth license plate correction module outputs the license plate correction information corresponding to the target vehicle and the confidence of the license plate correction information. Among them, the weight corresponding to the fourth license plate correction module is β4. The weight β4 corresponding to the fourth license plate correction module is obtained by pre-training the license plate correction model.

[0175] In one embodiment, based on the license plate detection information and vehicle detection information of the detected vehicle and the target vehicle respectively corresponding to the neighboring space-time corresponding to the preset area and the preset time period, the vehicle similarity between the detected vehicle and the target vehicle is determined. In response to the vehicle similarity reaching a fourth threshold, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity.

[0176] In one embodiment, the fifth license plate correction module can be used to correct the abnormal license plate based on the license plate detection information and vehicle detection information corresponding to the detection vehicle and the target vehicle respectively corresponding to the neighboring space-time corresponding to the preset area and the preset time period, and the fifth license plate correction module outputs the license plate correction information corresponding to the target vehicle and the confidence of the license plate correction information. Among them, the weight corresponding to the fifth license plate correction module is β5. The weight β5 corresponding to the fifth license plate correction module is obtained by pre-training the license plate correction model.

[0177] In one embodiment, based on the running trajectory of the target vehicle, the target vehicle's necessary trajectory points are searched in a preset map to determine; based on the license plate detection information and vehicle detection information corresponding to each detected vehicle and the target vehicle collected at the necessary trajectory points, the vehicle similarity between the detected vehicle and the target vehicle is determined; in response to the vehicle similarity reaching a fifth threshold, the detected license plate number and / or detected license plate type corresponding to the target vehicle is corrected based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity.

[0178] In one embodiment, the sixth license plate correction module can be used to search in a preset map based on the running trajectory of the target vehicle to determine the necessary trajectory points of the target vehicle; based on the license plate detection information and vehicle detection information corresponding to each detected vehicle and the target vehicle collected at the necessary trajectory points, the abnormal license plate is corrected, and the sixth license plate correction module outputs the license plate correction information corresponding to the target vehicle and the confidence of the license plate correction information. Among them, the weight corresponding to the sixth license plate correction module is β6. The weight β6 corresponding to the sixth license plate correction module is obtained by pre-training the license plate correction model.

[0179] In one embodiment, the first license plate correction module, the second license plate correction module, the third license plate correction module, the fourth license plate correction module, the fifth license plate correction module and the sixth license plate correction module respectively perform license plate detection based on the running track, vehicle detection information and license plate detection information of the target vehicle with abnormal license plate, and obtain the license plate correction information and license plate correction information confidence corresponding to each license plate correction module. The correction results of the weights of each license plate correction module, the license plate correction information and the license plate correction information confidence are fused to obtain the license plate correction result of the target vehicle.

[0180] In one embodiment, a detected vehicle can be selected as a target vehicle from the detected vehicles corresponding to the preset area and the preset time period, and the running track of the target vehicle can be displayed on the interactive interface. Based on the running track of the target vehicle, a candidate area and / or candidate must-pass track points for collecting the target vehicle can be outlined on the preset map. It is also possible to select only the captured images collected by the image acquisition device of part of the preset area for analysis. The preset number of correction results corresponding to the abnormal license plate are displayed through the license plate correction model, so that the correction results of the abnormal license plate can be displayed according to the actual situation.

[0181] The license plate correction method provided in this embodiment obtains the running track of the target vehicle within a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information; based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle, determines whether the license plate detection information of the target vehicle is an abnormal license plate; in response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected. In this application, it is determined whether the license plate detection information of the target vehicle is an abnormal license plate through the running track, license plate detection information and vehicle detection information of the target vehicle, and then the license plate detection information corresponding to the abnormal license plate is corrected to improve the accuracy of the vehicle license plate information.

[0182] See also Figure 2 , Figure 2 It is a schematic diagram of the framework of an embodiment of a license plate correction device provided by the present invention.

[0183] This embodiment provides a license plate correction device 60 , which includes an acquisition module 61 , an analysis module 62 and a correction module 63 .

[0184] The acquisition module 61 is used to acquire the running track of the target vehicle in a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information.

[0185] The analysis module 62 is used to determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle.

[0186] The correction module 63 is used for correcting the license plate detection information of the target vehicle in response to the license plate detection information of the target vehicle being an abnormal license plate.

[0187] In the license plate correction device provided in this embodiment, whether the license plate detection information of the target vehicle is an abnormal license plate is determined through the target vehicle's running trajectory, license plate detection information and vehicle detection information, and then the license plate detection information corresponding to the abnormal license plate is corrected to improve the accuracy of the vehicle license plate information.

[0188] See also Figure 3 , Figure 3 80 is a schematic diagram of a terminal embodiment provided by the present invention. The terminal 80 includes a memory 81 and a processor 82 coupled to each other, and the processor 82 is used to execute program instructions stored in the memory 81 to implement the steps of any of the above-mentioned license plate correction method embodiments. In a specific implementation scenario, the terminal 80 may include but is not limited to: a microcomputer, a server, and in addition, the terminal 80 may also include a mobile device such as a laptop computer and a tablet computer, which is not limited here.

[0189] Specifically, the processor 82 is used to control itself and the memory 81 to implement the steps of any of the above-mentioned license plate correction method embodiments. The processor 82 can also be called a CPU (Central Processing Unit). The processor 82 may be an integrated circuit chip with signal processing capabilities. The processor 82 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 82 can be implemented by an integrated circuit chip.

[0190] In the above scheme, the license plate correction method includes:

[0191] Obtain the running track of the target vehicle in a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information;

[0192] Determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle;

[0193] In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected.

[0194] See also Figure 4 , Figure 4 The computer-readable storage medium 90 stores program instructions 901 that can be executed by a processor, and the program instructions 901 are used to implement the steps of any of the above-mentioned license plate correction method embodiments.

[0195] In the above scheme, the license plate correction method includes:

[0196] Obtain the running track of the target vehicle in a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information;

[0197] Determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or license plate detection information and vehicle detection information corresponding to the target vehicle;

[0198] In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected.

[0199] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0200] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0201] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0202] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0203] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0204] The above are only implementation modes of the present invention, and are not intended to limit the patent protection scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A license plate correction method, characterized in that: The license plate correction method comprises: Obtaining the running track of a target vehicle within a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information; the target vehicle has a corresponding vehicle-related object; Determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle; In response to the license plate detection information of the target vehicle being an abnormal license plate, correcting the license plate detection information of the target vehicle; Wherein, determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle includes: Determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the vehicle-associated object corresponding to the target vehicle, the license plate detection information, and the vehicle detection information; Wherein, determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the vehicle-associated object corresponding to the target vehicle, the license plate detection information, and the vehicle detection information includes: Determine whether the vehicle-associated object corresponding to the target vehicle is consistent with the vehicle-associated objects corresponding to other detected vehicles except the target vehicle corresponding to the preset area and the preset time period; in response to the vehicle-associated object of the target vehicle being consistent with the vehicle-associated object of at least one of the detected vehicles, determine the vehicle similarity between the target vehicle and the detected vehicle corresponding to the same vehicle-associated object based on the license plate detection information and the vehicle detection information corresponding to the target vehicle and the detected vehicle respectively; in response to the vehicle similarity being greater than a first threshold, determine that the license plate detection information of the target vehicle is an abnormal license plate.

2. The license plate correction method according to claim 1, characterized in that: The step of obtaining the running track of the target vehicle in a preset area and a preset time period includes: Obtaining the running tracks of all the detection vehicles corresponding to the preset area and the preset time period, each of the detection vehicles has the corresponding license plate detection information and vehicle detection information; Each of the detected vehicles is sequentially used as the target vehicle.

3. The license plate correction method according to claim 1, characterized in that: The determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle includes: Using an abnormal license plate detection model to determine a license plate detection result of the license plate detection information of the target vehicle based on the running track corresponding to the target vehicle, the license plate detection information and the vehicle detection information; the license plate detection result is an abnormal license plate or a normal license plate; The abnormal license plate detection model includes multiple detection modules. The adopting of the abnormal license plate detection model to determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the running trajectory corresponding to the target vehicle, the license plate detection information and the vehicle detection information includes: Based on each of the detection modules, abnormal license plate detection is performed based on the running track corresponding to the target vehicle, the license plate detection information, and the vehicle detection information, to obtain initial detection results corresponding to each of the detection modules; The initial detection results corresponding to each detection module are weightedly summed to obtain the license plate detection result corresponding to the license plate detection information of the target vehicle.

4. The license plate correction method according to claim 1, characterized in that: The license plate detection information includes the detected license plate number and the detected license plate type; the running trajectory includes the trajectory points that the target vehicle has passed through successively; the target vehicle has a corresponding vehicle-related object; The determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle includes: Comparing the detected license plate number of the target vehicle with the string regularization rules of each preset license plate number in the database; in response to the detected license plate number not matching the string regularization rules of each preset license plate number in the database, determining that the license plate detection information of the target vehicle is an abnormal license plate; or, determining whether the detected license plate number corresponding to the target vehicle matches the detected license plate type; in response to the detected license plate number corresponding to the target vehicle not matching the detected license plate type, determining that the license plate detection information of the target vehicle is an abnormal license plate; Or, based on the spatiotemporal information of each of the track points in the running track of the target vehicle, determining the detection running speed and / or the detection time difference of the target vehicle; in response to the detection running speed and / or the detection time difference corresponding to the target vehicle not meeting the corresponding preset requirements, determining that the license plate detection information of the target vehicle is an abnormal license plate; Or, the license plate number of the target vehicle is identified by the first device and the second device respectively, to obtain two detection license plate numbers corresponding to the target vehicle; in response to the detection license plate numbers of the target vehicle respectively detected by the first device and the second device being inconsistent, it is determined that the license plate detection information of the target vehicle is an abnormal license plate; or, determining whether the target vehicle has the corresponding detection license plate number and the detection license plate type; in response to the target vehicle not having the corresponding detection license plate number or the detection license plate type, determining that the license plate detection information of the target vehicle is an abnormal license plate; Or, based on the running trajectory of the target vehicle, the number of times the target vehicle appears at each of the trajectory points within a preset sub-period is statistically obtained; the preset sub-period is included in the preset period; in response to the number of times the target vehicle appears at the trajectory point within the preset sub-period does not meet the preset number, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

5. The license plate correction method according to claim 1, characterized in that: The determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle includes: Compare the license plate detection information of the target vehicle with each preset license plate information in the database; In response to the license plate detection information of the target vehicle being different from each of the preset license plate information in the database, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

6. The license plate correction method according to claim 5, characterized in that: The determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle includes: In response to the license plate detection information of the target vehicle being identical to a preset license plate information in a database, comparing the vehicle detection information of the target vehicle with the preset vehicle information corresponding to the preset license plate information; In response to the vehicle detection information of the target vehicle being different from the preset vehicle information corresponding to the preset license plate information, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

7. The license plate correction method according to claim 2, characterized in that: The license plate detection information includes the detected license plate number; The determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the vehicle-associated object corresponding to the target vehicle, the license plate detection information, and the vehicle detection information includes: Based on the vehicle-related object of the target vehicle, searching a database for a preset license plate number corresponding to the vehicle-related object; In response to the detection license plate number corresponding to the vehicle-associated object of the target vehicle being inconsistent with the preset license plate number, determining the vehicle similarity between the detection license plate number and the preset license plate number based on the license plate detection information corresponding to the target vehicle, the vehicle detection information and the vehicle registration information corresponding to the preset license plate number; In response to the vehicle similarity being greater than a second threshold, it is determined that the license plate detection information of the target vehicle is an abnormal license plate.

8. The license plate correction method according to claim 1, characterized in that: In response to the license plate detection information of the target vehicle being an abnormal license plate, correcting the license plate detection information of the target vehicle includes: In response to the license plate detection information of the target vehicle being an abnormal license plate, the license plate detection information of the target vehicle is corrected using a license plate correction model to obtain a license plate correction result of the target vehicle; Wherein, the license plate correction model includes multiple license plate correction modules. The method of using the license plate correction model to correct the license plate detection information of the target vehicle to obtain the license plate correction result of the target vehicle includes: Based on each of the license plate correction modules, the license plate detection information of the target vehicle is corrected respectively to obtain the license plate correction information output by each of the license plate correction modules; The license plate correction information output by each of the license plate correction modules is weightedly summed to obtain a license plate correction result of the target vehicle.

9. The license plate correction method according to claim 2, characterized in that: The license plate detection information includes the detected license plate number and the detected license plate type, and both the target vehicle and the detected vehicle have corresponding vehicle-related objects; The license plate correction method comprises: Filter out the detected vehicles corresponding to abnormal license plates from all the detected vehicles corresponding to the preset area and the preset time period, and obtain a normal license plate data set corresponding to the preset area and the preset time period; the normal license plate data set includes the detected vehicles corresponding to normal license plates; The correcting the license plate detection information of the target vehicle includes: Based on the detected license plate numbers and the detected license plate types corresponding to each detected vehicle in the normal license plate data set, a first relationship table corresponding to license plate numbers and license plate types is obtained; the detected license plate number of the target vehicle is compared with each license plate number in the first relationship table; based on the license plate type corresponding to the license plate number of the same type as the detected license plate number, the detected license plate type of the target vehicle is corrected; Or, based on the license plate detection information and the vehicle associated object corresponding to each detected vehicle in the normal license plate data set, a second relationship table corresponding to the license plate information and the associated object is obtained; the vehicle associated object of the target vehicle is compared with each associated object in the second relationship table; based on the license plate information corresponding to the associated object that is the same as the vehicle associated object of the target vehicle, the detected license plate number and / or the detected license plate type corresponding to the target vehicle is corrected.

10. The license plate correction method according to claim 2, characterized in that: The license plate detection information includes the detected license plate number and the detected license plate type. The correcting the license plate detection information of the target vehicle includes: Compare the detected license plate number of the target vehicle with the preset license plate number in the third relationship table; the third relationship table is a pre-constructed correspondence table between the preset license plate number and the preset license plate type; based on the preset license plate type corresponding to the preset license plate number of the same type as the detected license plate number of the target vehicle, correct the detected license plate type of the target vehicle; or, based on the license plate detection information corresponding to the target vehicle and the license plate information and vehicle information corresponding to each registered vehicle in the database, determining the vehicle similarity between the target vehicle and the registered vehicle; in response to the vehicle similarity reaching a third threshold, correcting the detected license plate number and / or the detected license plate type corresponding to the target vehicle based on the license plate information of the registered vehicle corresponding to the vehicle similarity; Or, based on the license plate detection information and the vehicle detection information corresponding to the detected vehicle and the target vehicle respectively corresponding to the neighboring space-time of the preset area and the preset time period, determine the vehicle similarity between the detected vehicle and the target vehicle; in response to the vehicle similarity reaching a fourth threshold, correct the detected license plate number and / or the detected license plate type corresponding to the target vehicle based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity; Or, based on the running trajectory of the target vehicle, the target vehicle must pass through the preset map and the search is performed to determine the trajectory points that the target vehicle must pass through; based on the license plate detection information corresponding to each detected vehicle and the target vehicle collected at the necessary trajectory points and the vehicle detection information, the vehicle similarity between the detected vehicle and the target vehicle is determined; in response to the vehicle similarity reaching a fifth threshold, the detected license plate number and / or the detected license plate type corresponding to the target vehicle is corrected based on the license plate detection information of the detected vehicle corresponding to the vehicle similarity.

11. A license plate correction device, characterized in that: The license plate correction device comprises: An acquisition module is used to acquire the running track of a target vehicle within a preset area and a preset time period; the target vehicle has corresponding license plate detection information and vehicle detection information; the target vehicle has a corresponding vehicle-related object; An analysis module, for determining whether the license plate detection information of the target vehicle is an abnormal license plate based on the running track and / or the license plate detection information and the vehicle detection information corresponding to the target vehicle; wherein, it is also used to determine whether the license plate detection information of the target vehicle is an abnormal license plate based on the vehicle-associated object corresponding to the target vehicle, the license plate detection information and the vehicle detection information; for judging whether the vehicle-associated object corresponding to the target vehicle is consistent with the vehicle-associated objects corresponding to other detection vehicles other than the target vehicle corresponding to the preset area and the preset time period; in response to the vehicle-associated object of the target vehicle being consistent with the vehicle-associated object of at least one detection vehicle, determining the vehicle similarity between the target vehicle and the detection vehicle corresponding to the same vehicle-associated object based on the license plate detection information and the vehicle detection information respectively corresponding to the target vehicle and the detection vehicle; in response to the vehicle similarity being greater than a first threshold, determining that the license plate detection information of the target vehicle is an abnormal license plate; The correction module is used to correct the license plate detection information of the target vehicle in response to the license plate detection information of the target vehicle being an abnormal license plate.

12. A terminal, characterized in that: The terminal includes a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor is used to execute program data to implement the steps in the license plate correction method according to any one of claims 1 to 10.

13. 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 the processor, the steps in the license plate correction method according to any one of claims 1 to 10 are implemented.

Images