LogGabor和SVM算法在指纹分割中的应用

Application of LogGabor and SVM algorithm in fingerprint segmentation

针对目前指纹图像分割算法的不足,提出了一种基于Log-Gabor和SVM算法的指纹分割.该方法先将指纹进行Log-Gabor变换以提取特征构成特征向量,再利用SVM进行特征向量的分类处理,最后完成对指纹的分割.实验结果表明,与目前常用的指纹分割法相比,该算法能够更为高效合理地对指纹图像进行分割.

According to the shortage of present fingerprint segmentation algorithm, this paper puts forward fingerprint segmentation based on LogGabor and SVM algorithm. This method first carries out LogGabor transform of fingerprint to extract the features and form feature vector, then carries out classified treatment of feature vector using SVM, finally realizes segmenta tion of fingerprint. Experimental results show that this method can realize fingerprint segmenta tion more efficiently and reasonably compared to common method.