基于模板匹配的掌纹识别自适应阈值选取方法

Adaptive thresholding method for online palmprint image verification based on template matching

结合在线掌纹的特点,对模板匹配算法进行了改进.首先为了去除噪声,在匹配之前对掌纹图像做小波滤波处理;然后对匹配之后的数值矩阵采用各向异性的自适应阈值处理方法,以克服光照不均匀及光的方向性等问题,有效提取掌纹线特征;最后应用本算法对不同掌纹图像进行识别,验证了方法的有效性.

According to the characteristics of online palmprint, the template matching method is improved. At first, wavelet filter processing is adopted on palmprint image before the matching process in order to erase the noises. Then anisotropic adaptive thresholding method is applied to the numerical value matrix after the matching process to overcome the problems such as. the ununiformity of illumination and the directionality of light, and extract the line features of palmprint effectively. Finally, this method is applied to verify various palmprint images, which proves the effectiveness of the method.