Blurred license plate identification method and apparatus

Abstract

The present invention provides a blurred license plate identification method and apparatus. The method comprises: obtaining MxN blurred license plate images being subjected to the inclination correction processing; carrying out graying processing on the MxN blurred license plate images to obtain MxN blurred license plate grayscale images; carrying out vertical projection on the MxN blurred licenseplate images to obtain N vertical projection values, and removing the upper and lower boundaries of the MxN gray license plate grayscale images according to the N vertical projection values; carryingout scaling processing on the blurred license plate grayscale images being subjected to the upper and lower boundary removal processing; carrying out horizontal projection on the scaled blurred license plate grayscale images to obtain M horizontal projection values, and determining traversal points on the scaled blurred license plate grayscale images according to the M horizontal projection values; and using a character template generated based on the neural network to identify the traversal points on the scaled blurred license plate grayscale images, and outputting a blurred license plate identification result. According to the method and apparatus provided by the present invention, a blurred license plate can be accurately and efficiently identified.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and device for recognizing a blurred license plate. Background technique [0002] With the development of video surveillance technology, intelligent monitoring of traffic scenes has become an important part of realizing traffic intelligence. At present, in terms of intelligent monitoring of traffic scenes, the license plate number, as a unique identity feature of motor vehicles, is an extremely important traffic information, and license plate recognition is an indispensable function in intelligent transportation systems. The current license plate recognition method mainly includes three steps, namely license plate region location, license plate character segmentation and license plate character recognition. [0003] For the location of the license plate area, the license plate area is mainly extracted by using the color, texture, and edge information of the lic...

AI Technical Summary

Problems solved by technology

[0005] For the blurred license plate image, because the characters in the binarized image are connected together, or the upper, lower, left, and right boundaries of the license plate are connected with the characters, the binarized blurred license plate cannot be accurately boundary removed and character segmented. The characters of the fuzzy license plate will not be correctly recognized; in addition, because the existing technology is to identify all the traversal points, that is, blindly traverse the fuzzy license plate, this will not only make the recognition time longer, but also increase the error recognition rate accordingly

Images