License plate region positioning method and system based on video images

Abstract

The invention discloses a license plate region positioning method and system based on video images. The method comprises steps as follows: image longitudinal feature detection, image binaryzation, communication of transversely adjacent regions of the images, longitudinal cutting of the images, merging of related sub-regions and selection of blocks. The system comprises an image longitudinal feature detection module, an image binaryzation module, a transversely adjacent region communication module, a longitudinal cutting module, a related sub-region merging module and a block selection module. An edge detection algorithm adopted by the method and the system is very good in compatibility with license plate colors, image backgrounds and image quality, features of license plate regions can be effectively extracted for different colors of license plates and seriously fuzzy images, the vehicle video images with complex backgrounds can be processed, the license plate regions in the images can be positioned stably and efficiently, problems that existing algorithms are higher in computational complexity and take too long computation time are solved, and the accuracy of license plate positioning and recognition is effectively improved.

Description

technical field [0001] The invention relates to a method for locating a license plate, in particular to a method and system for locating a license plate area based on a video image. It belongs to the field of license plate recognition. Background technique [0002] With the continuous improvement of productivity and the increasing number of cars in various cities, the traditional traffic management mode is facing great challenges. In recent years, computer technology has been widely used in the field of transportation. Such as traffic system, vehicle annual inspection system, community automatic toll collection system, road checkpoint management system. All are transforming to automated management, and the automatic license plate recognition technology has begun to be recognized by all walks of life. Therefore, the video image automatic license plate recognition technology is an irreplaceable important function in the field of traffic management, and it has high economic ...

AI Technical Summary

Problems solved by technology

Both the edge-based and color-based license plate area positioning algorithms rely too much on the contrast between the license plate edge and the license plate. If the edge ratio of the image is blurred, it will seriously affect the license plate positioning effect. In addition, the edge detection algorithm is too lazy to image. Clarity; and the color-based algorithm has too high requirements on the color contrast and color accuracy of the image, and the positioning parameter deviation between the ring shadow illuminated by strong light and the dark surround view is too large, and the adaptability is not strong; while the algorithm based on The algorithm of machine learning is mainly realized by artificial neural network technology, and this type of algorithm can basically achieve a detection rate of more than 85% in the license plate area

However, the calculation complexity of these algorithms is high, and the calculation time required is too long, so it is difficult to adapt to the application environment with demanding calculation time such as high-speed functions.

Images