中文说明： 通过反复训练模板能有较高的识别率，主要是基于mtlab的程序，基于负熵最大的独立分量分析，基于欧几里得距离的聚类分析，应用小区域方差对比，程序简单，主要为数据分析和统计。
English Description：
Through practice templates can have a higher recognition rate is mainly based on the mtlab program, independent component analysis based on maximum negentropy, cluster analysis based on Euclidean distance, small regional variance correlation and simple program, primarily for data analysis and statistics.