Palmprint Recognition

Biometric technologies are becoming the foundation of new generational solutions for highly secure identification and personal verification. Palmprint recognition and verification are gaining popularity and diverse applications for the reason that they are considered to be non-invasive, low cost, and natural biometric technologies.This thesis studies palmprint recognition and develops a palmprint recognition and verification prototype system. The main contents of the thesis are listed as follows:1. Image preprocessing technologies for palmprint recognition are studied. The alignment of ROI(Region of Interest) and gray preprocessing are discussed. The alignment of ROI based on maximal inscribed circle is presented.2. Advantages and disadvantages of current palmprint recognition technologies are discussed. A new algorithm named Fisherpalm is proposed. And Gabor wavelet is selected to enhance the robust to lighting variations and efficiency. The principle of Fisherpalm algorithm is introduced in detail. Functions of the algorithm are tested using the proposed evaluation criterion. The experimental results indicate that the proposed algorithm provides better performance compared with Fisherpalm algorithm.3. Based on the requirement and functional analysis of the system, the workflow and functional modules of the system are made. The palmprint recognition and verification prototype system's schemes of the interface, enrollment, recognition and verification are designed and the key techniques for the system are analyzed.4. A palmprint recognition and verification prototype system is developed with functions of ROI alignment, sub-image preprocessed, features extracted, user enrollment, user recognition and user verification. The performance of the system is tested on public palmprint database(BAC2005 and PolyU_Palmprint_Database) and real-life application environment. The experimental results indicate that the system achieves high recognition rate (EER <5%) at real-time working environment.