A system and method for acquiring high point vehicle ar float labels based on linked low point pin data

Abstract

The present application belongs to the field of image recognition, high-precision AR augmented reality, high-low point linkage and the like, and particularly relates to a system and method for obtaining a high-point vehicle AR floating label based on linkage low-point portal data, wherein low-point portals are separately installed on each lane on the road, the low-point portals continuously identify vehicle license plates through image recognition, and generate floating AR labels, and the low-point portals of different lanes upload the identified vehicle license plates, vehicle types, vehicle colors and the like, as well as vehicle pictures, to a high-point camera background system. The detection ranges of the low-point portals at different positions are divided by the background system, and the vehicle queue in the region within the high-point line-of-sight range is matched with the corresponding license plate number according to the time stamp, vehicle type and vehicle color.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of image recognition, high-precision AR augmented reality, high-low point linkage, and specifically relates to a system and method for obtaining high-point vehicle AR floating labels based on linkage low-point portal data. BACKGROUND

[0002] Currently, due to the limited resolution of high-point cameras and the long time required for long-distance zoom positioning, multiple vehicle license plates cannot be recognized at the same time, so high-point cameras can generally only recognize coarse-grained information such as vehicle images, sizes, and colors. Therefore, high-point cameras generally need vehicles to be equipped with communication intelligent devices for data uploading.

[0003] Currently, low-point portals generally have image recognition algorithms for recognizing license plates, vehicle types, and vehicle colors, with recognition speeds of less than 100 ms. However, these information is not fully utilized, and high points generally need to be equipped with communication intelligent devices to obtain AR detailed information. Installing devices requires high equipment costs, is difficult to popularize and spread, and is prone to incomplete vehicle information statistics, which cannot effectively solve the problem of high-point camera efficient acquisition of AR license plate labels. SUMMARY

[0004] To solve the above technical problems, the application provides a system and method for obtaining high-point vehicle AR floating labels based on linkage low-point portal data.

[0005] The application is implemented by the following technical solutions:

[0006] A method for obtaining high-point vehicle AR floating labels based on linkage low-point portal data, comprising the following steps:

[0007] Identify the first lane structured data, and the first camera system identifies the structured data and vehicle pictures of the vehicles on the first lane through image recognition technology;

[0008] Send the structured data and vehicle pictures to the second camera system;

[0009] Obtain the vehicle information sequence, and the second camera system identifies the vehicle features in the video captured by the second camera, sorts the vehicle features according to the driving direction of the lane, and obtains the vehicle information sequence;

[0010] The AR label is matched, the acquired structured data of the vehicle on the first lane is sorted according to a timestamp to obtain a structured data sequence of the first lane, and is matched with the vehicle information sequence; when the structured data sequence of the first lane matches a certain piece of data in the vehicle information sequence, a vehicle pixel matrix in a video captured by the second camera is acquired and matched with the license plate AR label in the structured data of the vehicle on the first lane.

[0011] The AR label is bound, a video picture of the second camera is acquired, and the AR label is displayed in the video picture and bound with the vehicle pixel matrix.

[0012] Further, the structured data comprises a vehicle license plate, a vehicle type and a vehicle color.

[0013] Further, the vehicle features comprise a vehicle type and a vehicle color.

[0014] Further, the vehicle type is classified according to vehicle size, and is classified into a small vehicle, a medium vehicle, a large vehicle and a giant vehicle.

[0015] Further, the high-point video picture continuously performs image recognition, distinguishes a single vehicle pixel matrix, and performs subtraction calculation on the vehicle pixel matrices of two frames of video pictures in front and back, and when the calculated difference is within a certain range, the two pixel matrices are judged as the same vehicle image, and the same AR label is used.

[0016] The application also relates to a system for acquiring high-point vehicle AR floating labels based on linkage low-point aperture data, comprising a low-point aperture camera, a high-point camera and a high-point camera background system; wherein,

[0017] The low-point aperture camera is used for identifying structured data and vehicle pictures of vehicles on a first lane through image recognition technology, and uploading the structured data and vehicle pictures to the high-point camera background system through a fiber or wireless communication module;

[0018] The high-point camera background system identifies vehicle features in a video captured by the high-point camera, sorts the vehicle features according to a lane driving direction to obtain a vehicle information sequence, sorts acquired structured data of vehicles on the first lane according to a timestamp to obtain a structured data sequence of the first lane, and matches the structured data sequence with the vehicle information sequence; when the structured data sequence of the first lane matches a certain piece of data in the vehicle information sequence, a vehicle pixel matrix in a video captured by the high-point camera is acquired and matched with the license plate AR label in the structured data of the vehicle on the first lane.

[0019] The high point camera is used to acquire a high point video picture and display the AR label in the video picture, and bind the vehicle pixel matrix.

[0020] Further, the structured data includes a vehicle license plate, a vehicle type and a vehicle color.

[0021] Further, the vehicle features include a vehicle type and a vehicle color.

[0022] Further, the vehicle type is classified according to vehicle size, including small cars, medium cars, large cars and giant cars.

[0023] Further, the high point video picture continuously performs image recognition to distinguish the pixel matrix of a single vehicle, and subtracts the vehicle pixel matrix of the front and rear frames of the video picture to calculate the difference, and when the difference is within a certain range, the two pixel matrices are determined as the same vehicle image, and the same AR label is used.

[0024] The present application provides a system and method for obtaining high point vehicle AR floating labels based on linked low point portal data, which innovatively uploads low point portal recognized vehicle data to high point video for analysis and full utilization, supplements high point video information, and reduces information acquisition cost. At the same time, video picture continuous comparison is used to determine the moving track of the same object to bind the AR license plate label. Other vehicle details can also be bound in this way. Make full use of existing low point portals, use emerging technologies such as AR technology, image recognition technology and video analysis technology, obtain a large amount of detailed information of vehicles at low cost at high points, and bind AR labels with vehicle pixel matrices, so that high point video obtains AR license plate labels of global vehicles at low cost, and other vehicle information can also be obtained in this way. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0026] In one embodiment, a system for obtaining high point vehicle AR floating labels based on linked low point portal data is provided, and the required devices include a low point portal camera, a high point camera and a high point camera background system.

[0027] Low point pinhole camera: through image recognition technology, quickly identify the vehicle license plate, vehicle type, vehicle color and other structured data and vehicle pictures on a specific lane, and upload the structured data and vehicle pictures to the high point camera background system through the optical fiber or wireless communication module.

[0028] High point camera background system: the background system identifies the vehicle features in each low point pinhole detection range through video recognition, sorts the vehicle types and colors according to the driving direction of the lane, and classifies the vehicle types according to the vehicle size, including small cars, medium cars, large cars, and giant cars. The background system sorts the vehicle structured data received from a specific lane according to the timestamp, and matches it with the vehicle information sequence identified by the image recognition at the same timestamp. When the structured data uploaded by the low point pinhole matches a certain piece of data in the vehicle sequence parsed from the video, the vehicle pixel matrix in the video can be bound with the license plate AR tag.

[0029] High point camera: obtain the video picture of the high point, and fuse the AR tag generated by the background system with the video picture.

[0030] The application also provides a method for obtaining high point vehicle AR floating tags based on linked low point pinhole data, which is applied to the above-mentioned system. It includes: installing low point pinholes for each lane on the road, quickly identifying the vehicle license plate, vehicle type, vehicle color and other structured data on a specific lane through image recognition technology, and uploading the structured data to the high point camera background system through the optical fiber or wireless communication module.

[0031] The high point camera background system divides the detection range of different low point pinhole cameras, and the background system identifies the vehicle features in each low point pinhole detection range through video recognition, sorts the vehicle types and colors according to the driving direction of the lane, and classifies the vehicle types according to the vehicle size, including small cars, medium cars, large cars, and giant cars.

[0032] The background system sorts the vehicle structured data received from a specific lane according to the timestamp, and matches it with the vehicle information sequence identified by the image recognition at the same timestamp. When the structured data uploaded by the low point pinhole matches a certain piece of data in the vehicle sequence parsed from the video, the vehicle pixel matrix in the video can be bound with the corresponding AR license plate tag.

[0033] For example: two red large vehicles appear in the lane, followed by a white small vehicle, and then a blue medium-sized vehicle. The color and type of the vehicle have a certain probability of repetition, but the order of the vehicle combination is indeed difficult to repeat and continues to extend. After the order combination of the vehicle color type, a certain group of vehicles in the low-point portal detection range can be quickly matched, and the specific license plate of the adjacent vehicle of the same color and type can be distinguished according to the time stamp of the license plate. The vehicle with an earlier time stamp must be the first to enter the portal detection range according to the lane driving direction in the combination, so the license plate can be assigned.

[0034] The high-point video picture continuously performs image recognition, distinguishes the pixel matrix of a single vehicle, and subtracts the vehicle pixel matrix of the front and rear two frames of video pictures. When the value is within a certain range, it is judged that the two pixel matrices are the same vehicle image, and the same AR vehicle label is used. In this way, the label can move with the pixel array to generate a high-point AR floating label.

[0035] The application also provides a computer readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the method for obtaining a high-point vehicle AR floating label based on linkage low-point portal data.

[0036] The application also provides a computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for obtaining a high-point vehicle AR floating label based on linkage low-point portal data when executing the program.

[0037] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the application. It should be understood that the above description is only a specific embodiment of the application and does not limit the protection scope of the application. Any modification, equivalent replacement, improvement, etc. without departing from the spirit and scope of the application also belongs to the protection scope of the application.

Claims

1. A method for acquiring high point vehicle AR floating labels based on linked low point pin data, characterized in that, The method comprises the following steps: identifying first lane structured data, a first camera system identifying structured data and vehicle pictures of vehicles on the first lane through image recognition technology, the first camera system being a low-point camera; sending the structured data and vehicle pictures to a second camera system, the second camera system being a high-point camera; obtaining a vehicle information sequence, the second camera system identifying vehicle features in a video captured by the second camera, sorting the vehicle features according to the driving direction of the lane to obtain the vehicle information sequence; matching AR labels, sorting the obtained structured data of vehicles on the first lane according to timestamps to obtain a structured data sequence of the first lane, and matching the structured data sequence with the vehicle information sequence, when the structured data sequence of the first lane matches a certain piece of data in the vehicle information sequence, obtaining vehicle pixel matrices in the video captured by the second camera and license plate AR labels in the structured data of vehicles on the first lane; binding AR labels, obtaining a video picture of the second camera, and displaying the AR labels in the video picture and binding the AR labels with the vehicle pixel matrices; the structured data comprises vehicle license plates, vehicle types and vehicle colors; the method for obtaining high-point vehicle AR floating labels based on linked low-point camera data further comprises: continuously performing image recognition on a high-point video picture, distinguishing single vehicle pixel matrices, and subtracting vehicle pixel matrices in front and back frames of the video picture, when the calculated difference is within a certain range, judging that the two pixel matrices are the same vehicle image, and using the same AR label.

2. The method for acquiring high point vehicle AR floating labels based on linked low point socket data according to claim 1, wherein, the vehicle features comprise vehicle types and vehicle colors.

3. The method for acquiring high point vehicle AR floating labels based on linked low point socket data according to claim 2, characterized in that, the vehicle types are classified according to vehicle sizes, and are divided into small cars, medium cars, large cars and giant cars.

4. A system for acquiring high point vehicle AR floating labels based on linked low point pin data, the system comprising: It comprises: a low-point camera, a high-point camera and a high-point camera background system; wherein, the low-point camera is used to identify structured data and vehicle pictures of vehicles on the first lane through image recognition technology, and upload the structured data and vehicle pictures to the high-point camera background system through a fiber or wireless communication module; the high-point camera background system identifies vehicle features in a video captured by the high-point camera, sorts the vehicle features according to the driving direction of the lane to obtain a vehicle information sequence, sorts the obtained structured data of vehicles on the first lane according to timestamps to obtain a structured data sequence of the first lane, and matches the structured data sequence with the vehicle information sequence, when the structured data sequence of the first lane matches a certain piece of data in the vehicle information sequence, obtains vehicle pixel matrices in the video captured by the high-point camera and license plate AR labels in the structured data of vehicles on the first lane; the high-point camera is used to obtain a video picture of the high-point camera, and display the AR labels in the video picture and bind the AR labels with the vehicle pixel matrices; the structured data comprises vehicle license plates, vehicle types and vehicle colors; The method for obtaining the high-point vehicle AR floating label based on the linkage low-point bayonet data further comprises: continuously performing image recognition on the high-point video picture, distinguishing the pixel matrix of a single vehicle, and subtracting the vehicle pixel matrix of the front and rear two frames of the video picture, and when the calculated difference is within a certain range, judging that the two pixel matrices are the same vehicle image, and using the same AR label.

5. The system for acquiring high point vehicle AR floating labels based on linked low point keyhole data of claim 4, wherein, The vehicle features include a vehicle type and a vehicle color.

6. The system for acquiring high point vehicle AR floating labels based on linked low point keyhole data of claim 5, wherein, The vehicle type is classified according to vehicle size, and is classified into a small vehicle, a medium vehicle, a large vehicle and a giant vehicle.