非接触指纹图像分割与增强方法的研究

非接触指纹图像的背景区域比接触式的相对复杂,且脊、谷线的对比度也比较低,用一般的接触式指纹图像处理方法很难达到很好的处理效果。本文针对非接触指纹图像的这两个特点提出了合理的分割与增强方法。采用基于照明-反射模型的同态滤波思想进行指纹图像与背景区域的分割。在对指纹图像增强时首先采用相干滤波技术进行滤波增强,再用简化的Gabor函数模板进行二次增强。实验结果表明,本文提出的方法在一定条件下能够很好的实现指纹图像的背景分割,同时本文方法的增强效果相比于以往的也有了明显的提高。

Research of segmentation and enhancement methods for contactless fingerprint images

The background areas of contactless fingerprints are more complex than those of the contact. What＇s more, the contrast of the ridge and valley lines are much lower. So the general methods for contact fingerprint images are difficult to achieve good effect, In this paper, a method is put out to segment and enhance the images reasonably aiming at these two features The homomorphic filtering idea based on the illumination-reflectance mode is adopted to segment the fingerprint images and background regions. When the fingerprint images are enhanced, the coherence filtering technique is adopted firstly and then the simplified Gabor function template is used to enhance them again. Experiment results show that the proposed method can achieve the background segmentation of the fingerprint well under certain conditions, while the enhancement effect is improved significantly compared with those of the previous methods.