Research on Algorithm of Locating License Plate and Problem of Redundant Trailer Image in Real Traffic

Abstract

Locating license plate area in an image is crucial to traffic recognition. By analyzing several common algorithms and their traits, this paper establishes the improved locating algorithm, which is combined algorithm based on color features with edge features, adding neural network training model for classifying plate images. Through the algorithm test it shows that the method has a nice effect. At last, this paper proposes a problem of redundant trailer image happened in real traffic and point out a solution idea.