Description: 本文将小波图像分解和信息熵特征提取相结合，提出一种新的掌纹特征提取算法。该算法首先对掌纹灰度图像进行二维小波分解，再利用多分辨信息熵分别计算不同尺度下的能谱熵作为特征向量，从而实现掌纹特征提取。该算法不但避免了图像增强和纹理细化等预处理过程，而且运用多分辨信息熵的自适应计算方法来调节分解级数，使得到的特征向量长度远小于传统算法。-In this paper, wavelet image decomposition and information entropy feature extraction combined propose a new feature extraction algorithm. The algorithm first palmprint grayscale image two-dimensional wavelet decomposition, reuse multiresolution information entropy were calculated at different scales of the energy spectrum entropy as feature vectors, enabling feature extraction. The algorithm not only avoids the preprocessing such as image enhancement and texture refinement process, and the use of multi-resolution information entropy adaptive calculation method to regulate the decomposition level, so that the length of the feature vector obtained is far smaller than the conventional algorithm. Platform: |
Size: 8443904 |
Author:wei |
Hits:
Description: 本文将小波图像分解和信息熵特征提取相结合，提出一种新的掌纹特征提取算法。该算法首先对掌纹灰度图像进行二维小波分解，再利用多分辨信息熵分别计算不同尺度下的能谱熵作为特征向量，从而实现掌纹特征提取。该算法不但避免了图像增强和纹理细化等预处理过程，而且运用多分辨信息熵的自适应计算方法来调节分解级数，使得到的特征向量长度远小于传统算法。-In this paper, wavelet image decomposition and information entropy feature extraction combined propose a new feature extraction algorithm. The algorithm first palmprint grayscale image two-dimensional wavelet decomposition, reuse multiresolution information entropy were calculated at different scales of the energy spectrum entropy as feature vectors, enabling feature extraction. The algorithm not only avoids the preprocessing such as image enhancement and texture refinement process, and the use of multi-resolution information entropy adaptive calculation method to regulate the decomposition level, so that the length of the feature vector obtained is far smaller than the conventional algorithm. Platform: |
Size: 1056768 |
Author:wei |
Hits:
Search - preprocessing in palmprint