基于小波域软阈值的指纹分割方法研究

鉴于指纹图像的质量对于其后续识别步骤具有非常重要的影响,因而需要对采集的指纹图像进行增强处理.提出了一种新的基于小波域软阈值Contourlet变换的指纹图像分割算法.Contourlet变换是一种新的多尺度几何分析方法,它不仅拥有小波变换的时频局域特性和多分辨率特性,而且还具有很好的方向性和各向异性.实验结果表明,该算法能够有效提升指纹图像的质量,提高指纹纹线谷脊清晰度,且具有较好的连续性.

A fingerprint image segmentation based on soft-threshold wavelet domain

The quality of fingerprint image has great affect for its follow-up steps such as recognition.Therefore,an effective method is needed to enhance the collected fingerprint image.In this paper,a new fingerprint image enhancement approach based on Contourlet transform soft-threshold wavelet domain was proposed.Contourlet is a new multi-scale geometric analysis method,which not only has the main features of local time-frequency and multi-resolution,but also has the directionality and anisotropy.Experimental results demonstrated that the method could effectively improve the quality of fingerprint image,and could increase the clarity of fingerprint ridges and valleys,and have better continuity.