Palmprint recognition

• Main Author: Zheng, Liren.
• Other Authors: Tan Boon Tiong
• Format: Final Year Project
• Language: English
• Published: 2011
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/44441
• Institution: Nanyang Technological University

As global concerns for security continue to rise amidst technological advancement, biometric identification has become an integral part of high level security systems. There are many forms of biometric identification and the aim of this final year project is to implement a practical and secure palmprint recognition system that would form the basis for further development into a security system. The degree of accuracy of the system will also be investigated. Palmprint images were obtained using a scanner before the images are binarized. Thereafter, the area of interest of the palmprint would be located and cropped out. The cropped area will undergo Gabor filtering into the real, imaginary and orientation states before being used for matching with other processed palmprint images. The method to be used for matching one palmprint image to another would be the modified Hamming Distance formula. Upon comparison, a score of between 0 and 1 would be shown. Scores closer to 0 would denote that both palmprint images have high level of similarity and scores closer to 1 would denote that the images are very different. The results would be collected and tabulated and conclusions would be drawn as to whether this palmprint recognition system is accurate and in the event that it did not perform as well as expected, to identify the range of scores for accuracy of the system.