| With the rapid development of information technology and internet all over the world, information security is extraordinarily important, especially in personal identification technology. Biometrics technology is gaining more and more attention recently, because biological feature has unique stability, uniqueness and convenience. As a new biometric technology, palmprint recognition has been focused on in these years due to its advantages such as easy capturing, abundant image information, high-rate acceptance by users, difficulty in faking and fewer chances to be interfered by noise etc.. The thesis mainly investigates in palmprint recognition feature extraction. The main contents and originalities in this thesis are listed as follows:1,Three algorithms about locating and segmenting palmprint image into region-of-interest are analyzed and their performances are compared in computer simulations, in which the algorithm based on morphologic operator is used.2,A novel palmprint recognition method using Gabor wavelets and 2-Dimensional principal component analysis (2-DPCA) and PCA is used further feature dimensional reduction. Simulation results show that the proposed method is superior to that of Gabor plus PCA method in palmprint recognition and identification.3,An optimal 2-D orthogonal Log-Gabor filter is constructed by two 2-D Log-Gabor filters with orthogonal orientations. The filter is then used to extract orthogonal phase features from palmprint images. Hamming distance is used as a classifier in this method. Experimental results showed that the performance of this novel Log-Gabor filter is much superior to those of algorithms based on 2-D Gabor filters.4,A new 2-D mixture Log-Gabor filter is constructed by a number of weighted 2-D Log-Gabor filters with different orientations. This new filter is used to extract mixture phase feature from palmprint images. Experimental results show that correct recognition rates are further improved by using the proposed filter. |
PDF Full Text Request