Online Palmprint Identification Algorithm And Applications

The rapid development of information technology has promoted the progress of society. The functioning of today's society is increasingly depended on the information technology, which has put forward high requirements to the security of information and systems. It is a necessity to continuously strengthen the security mechanisms protecting for some confidential information in people's daily life, business, government and the army. Traditional security and authentication can not guarantee the safety of confidential information. However, Biometrics based on personal biologic features can solve the authentication problems efficiently and safely, which is concerned widely. Compared with other biometrics technology, palmprint recognition has been research later, but it has some unique advantages: easy to capture; low-resolution images can be employed; easy to be accepted; easy to fuse with other biometrics to recognition, etc. It is for those reasons that palmprint recognition has recently attracted an increasing amount of attention from researchers.Through the study and research to the development history and current research to this subject, we know that this technology is still in a stage of rapid development. On the basis of learning the main technologies of function modules respectively in palmprint recognition field, we mainly focused on the palmprint feature extraction technology. And then we proposed a novel palmprint feature extraction method based on both the Gabor filter bank and another texture operator: Local Binary Patterns (LBP). LBP can represent the local neighborhood relation information of texture, which is robust to the illumination variations, but is in single scale and sensitive to noise, while the Gabor filters can represent the texture information in several scales and orientations. Combining the advantages of both texture operator tools, we proposed a novel palmprint representation method, which brought higher recognition rate. We have made comparisons of palmprint recognition performance between other current main palmprint feature extraction methods and our proposed approach. The results are given in this paper. Finally, we also studied in detail the design of the various functional modules of a practical on-line palmprint identification system, and based on the C# platform, we developed a demonstration one, which includes the two main function modules: registration and recognition. However, instead of designing a picture achieving equipment, on the palmprint achieving step we loaded the already saved images from the public Hong Kong Polytechnic University's palmprint database. Based on these images, we finished the realization of other palmprint recognition modules.