Edge extraction of face feature based on wavelet transform

人脸图象识别是模式识别、计算机视觉和人工智能领域的一个热点 ,而人脸图象的特征提取又是人脸图象识别的关键步骤。小波变换以其多分辨率分析和多通道分解的独特分析方法在图象处理领域中有着广泛的应用。文章利用二进小波变换多尺度的特性 ,结合一些常规的边缘提取算法 ,在人脸图象特征的边缘提取方面取得了令人满意的结果 ,实验表明是一种有效的人脸图象特征边缘提取算法Human face recognition is a hot point in Pattern Recognition, Computer Vision and AI, and face feature extraction is a key part in face recognition. Wavelet transform (WT) is widely used in image processing for its special virtues of multi definition and multi channel analysis. A new approach based on WT is presented. Combined with some edge detection algorithm,the edge feature of human face image is obtained.Experiments show that the result is satisfying and this algorithm is successful in face image edge extraction