摘要： 设计了一个人脸画像自动识别系统,利用离散余弦变换进行特征提取,结合图像整体信息和局部信息,然后用最近邻分类器进行分类识别。实验获得了94.3%的正确识别率。

关键词: 人脸识别, 画像识别, 离散余弦变换

Abstract: This paper presents an automatic face painting recognition system. It adopts a DCT-based feature extraction algorithm, and combines the global information and the local information of the apparatus of the face. Finally a nearest neighbor classifier (NNC) is used to classify the face painting patterns. The experiment achieves a correct recognition accuracy of 94.3%.

中图分类号: 

• TP391.41