Research on the extraction to the Region of Interest area in palmprint

Abstract

Biometrics recognition is a kind of identification technology that uses the human's special physiology or behavior characteristic, it provides a kind of high reliability, good stability approach of identity verification. Palm print identification mainly formed by palm acquisition, preprocessing, palmprint feature extraction and matching modules. Preprocessing is to locate the image on the palm and normalized it, the morphology-based positioning method can quickly and accurately positioning palm print image, to partition ROI region and the main secondary attainment areas by the identification can improve the matching rate to overcome some of the palm print image defects; palm prints of different images within the same location of relatively fixed center cut, positioning can be used as a reference match, based on the maximum inscribed circle of the positioning method is simple, robust, less susceptible to image quality. The basic task of feature extraction is to study how to find the features on the classification and recognize the most effective features, Fisher's method compared to other methods of charge method can effectively extract the palm print features with a strong distinguished ability.