Offline Handwritten Chinese Character Recognition Based On PSO And SVM

As a branch of the character recognition technology, Off-line handwritten Chinese character recognition has become a major aspect of pattern recognition research field,its research is of great significance. Because off-line handwritten Chinese character recognition problem belongs to the classification problem of multi-class and complex pattern, high recognition rate need the integration of multi-classifiers through the former research, in the mean time, systems pending also increase more. Support Vector Machine(SVM)theory is a machine learning method developing from the basic of statistic learning theory, which shows special superiorities in dealing problems of small sample, nonlinear and multidimensional recognitions.It can get a biggest theory meaning and practice value that SVM theory is used for off-line handwritten Chinese character recognition. However, SVM has many disadvantages, notably the absence of fixed standardizations in the selection of kernel functions and parameters of SVM, only dealing with the two-sample problem and doing nothing for the multi-classification problem. Particle swarm optimization (PSO) is a stochastic optimization algorithm based on swarm intelligence principles. The method of optimized SVM parameter by PSO algorithm was presented, so a big difficulty of SVM parameter selection has been solved.Based on the traditional SVM, the paper presents PSO-SVMs combining PSO algorithm and directed acyclic graph. PSO-SVMs can quickly find the optimal SVM parameters which enhances the accuracy of classification and improves lack of two types of recognition of tradition SVM. At last, the experiment shows that improved SVMs for off-line handwriting recognition is of great feasiibility and effectiveness. The PSO-SVMs has an obvious advantage for the recognition function and results, there is a good improvement of traditional SVM.