License plate character recognition method based on SIFT operator and chaos genetic algorithm

Abstract

The invention belongs to a license plate recognition system, and discloses a license plate character recognition method based on an SIFT operator and a chaos genetic algorithm. The method is characterized in that Chinese characters and alphanumeric characters of a license plate are separately recognized, that is, the Chinese characters are recognized by using an SIFT operator feature extraction and template matching method; and the alphanumeric characters are recognized by using a thirteen-point feature extraction method and a support vector machine, and the problem of low overall recognition rate of the license plate characters due to that most of the existing license plate recognition systems adopt a unified character feature extraction and recognition method for recognition is solved. Meanwhile, in order to improve the classification capability of the support vector machine, the method adopts the chaos genetic algorithm to optimize radial basis function parameters and penalty factors, license plate images of different backgrounds are collected and tested and simulated on matlab software, the overall recognition rate of the characters can reach more than 99%, and the chaos genetic algorithm has a higher character recognition rate and a faster convergence rate than a traditional genetic algorithm.

Description

technical field [0001] The invention belongs to a license plate recognition system, and is a license plate character recognition method based on SIFT operator and chaotic genetic algorithm. Background technique [0002] The license plate recognition system has become the main means of intelligent traffic management in various occasions such as capturing violations, highway tolls, parking lot entrance and exit monitoring, vehicle speed and flow control, etc. Driven by the national economy, the automobile industry is booming, and the number of vehicles has also increased rapidly, but it has brought increasingly serious traffic problems, such as speeding, illegal parking, running red lights, traffic accidents, and chaotic fees. Therefore, the license plate recognition system is also Continuous innovation and improvement are required. In 2012, Guo Jinzhi proposed a license plate recognition system based on the SIFT algorithm in the publishing unit of Xidian University, which us...

AI Technical Summary

Problems solved by technology

In 2012, Guo Jinzhi proposed a license plate recognition system based on the SIFT algorithm in the publishing unit of Xidian University, which used the SIFT operator to extract features and template matching to recognize the characters of the license plate. The difference in structural complexity and local feature richness leads to a significant difference in the recognition rate between the two. As a result, the recognition rate of Chinese characters is high, and the recognition rate of numbers and letters is obviously low, and the overall recognition rate is not high.

Images