License plate recognition method and device

Abstract

The invention discloses a license plate recognition method and device. The method comprises the following steps: receiving a collected image sequence; performing vehicle detection on the image sequence to obtain a vehicle detection result and RE-ID features; performing vehicle tracking according to the vehicle detection result and the RE-ID features to obtain a tracking trajectory of the vehicle; performing license plate detection on the vehicle based on the tracking trajectory to obtain license plate characters; recognizing the license plate characters to obtain a current license plate recognition result of the vehicle; fusing the current license plate recognition result of the vehicle with the historical license plate recognition result of the vehicle to obtain a fusion result; judging whether the fusion result meets a set requirement or not; if yes, outputting the fusion result as a final license plate recognition result; otherwise, continuing to perform license plate detection and license plate recognition on the vehicle. According to the invention, the accuracy of the license plate recognition result can be improved.

Description

technical field [0001] The invention relates to the technical field of target recognition, in particular to a method and device for recognizing a license plate. Background technique [0002] License plate recognition has a wide range of applications in parking management, road traffic, etc., including vehicle detection, license plate text detection and correction, license plate text recognition, vehicle tracking and information fusion. [0003] A license plate recognition method provided by the prior art is as follows: figure 1 As shown, after the vehicle detection, the RE-ID (Re-Identification, re-identification) model is used to extract features for each vehicle, and then tracked by Deep Sort, and the license plate detection and license plate frame are performed on the image containing the vehicle. For the regression of the corner points, after using the positions of the four corner points for projection transformation and alignment, the license plate is scaled to a size ...

AI Technical Summary

Problems solved by technology

[0005] 1. With the increase of vehicles in the picture, the time consumption of extracting RE-ID features increases linearly, the tracking speed is slow, and the scene of fast license plate recognition cannot be handled

[0006] 2. The regression of the four corner points of the license plate, due to the difference in the installation of the license plate, leads to a large point error, which easily leads to the failure of the license plate correction

[0007] 3. For LPRNet whose license plate is scaled to 94x24, since the vertical scale is 24, the recognition rate of Chinese characters in provinces where the first Chinese character of the license plate is blurred or complex is low

Images