License plate character classification method

Abstract

The invention discloses a license plate character classification method, and relates to the field of character classification. The method comprises the following steps: S1, obtaining a training set dimension reduction matrix under each dimension reduction dimension through an SVD algorithm; obtaining a testing set dimension reduction matrix under each dimension reduction dimension based on the process of obtaining each training set dimension reduction matrix; S2, carrying out classification training on vectors of the dimension reduction matrix of each training set through an SVM algorithm to obtain a multi-class classifier of license plate characters under each dimension reduction dimension; S3, determining an optimal multi-class classifier; S4, converting a license plate character image to be recognized into a to-be-recognized dimension reduction vector, and predicting a character category to which the vector belongs by using the optimal multi-class classifier. The SVD algorithm reduces the dimension of the original matrix of training sets, and simplifies the parameters required for obtaining various multi-class classifiers; the time cost and the space cost of the obtained optimal multi-class classifier are low, the classification precision is high, and the method can be widely applied to embedded equipment with limited resources.

Description

technical field [0001] The invention relates to the field of character classification, in particular to a method for character classification of license plates. Background technique [0002] In the process of urban development, the number of automobiles continues to increase. With the popularization of artificial intelligence technology, automation management technology has penetrated into automobile management and other related industries. To realize the automated management of cars, it is first necessary to identify the identity "ID" of each car, which is the license plate. [0003] Currently popular character recognition methods are as follows: 1. Classification method based on template matching; 2. Character classification method based on neural network; 3. Character classification method based on SVM. [0004] Among them, the classification method based on template matching is low in cost and fast in recognition speed, and can basically meet the real-time requirements....

AI Technical Summary

Problems solved by technology

The implementation of the SVM-based character classification method is relatively simple, and it can achieve high classification accuracy without considering the training cost. However, the existing schemes mostly use the original character image data, and its calculation dimension is high, which makes the SVM model more complicated. , the amount of data storage is large, which increases the time cost of data classification

Images