The Study Palmprint Recognition Based On Gabor Filter

Palmprint recognition technology, as a relatively new biometric technology, has its own unique characteristics. A palmprint image includes many features, such as the main line, fold, back, detail points, singular points and texture. In the low-resolution palmprint images, lines and texture features contain the most and the clearest features. And the method based on direction is very effective because the direction of palm lines carries a lot of messages used to identify, and they is robust for the illumination. Compared with other biological characteristics, the palmprint image has many advantages. In the feature extraction, the palm provides a lot of messages to extract, and data collection is simple. People are very keen study of palmprint recognition, because of its low-cost acquisition devices, fast execution speed, high recognition rate and so on.However, the palmprint recognition technology is still in the development stage, in which theoretical and applied researches need to be further deepened and improved. Therefore, the in-depth study of palmprint recognition technology is very necessary. Given the successful application of Gabor filter in the pattern recognition, we will be introduced into the palmprint recognition, and do the corresponding researches.This paper first make brief of introduction about Gabor filters and corresponding methods, including Palm code, Fusion code, Competitive code and so on. And we propose a method of fast palmprint retrieval to use of Gabor energy information, which can effectively increase the recognition speed.Further, we simplify the Gabor filters and propose a fast algorithm for extracting features using the Simplified Gabor (SG) for the original methods based on Gabor filers. In these mothods, feature extraction is very time-consuming that using convolution and the original Gabor filter. We use a simplified version of Gabor filters to replace original, and the feature extraction speed greatly enhanced. Based on Gabor Phase recognition methods are used in iris and face recognition. In this paper, we use this Gabor Phase-based identification method in palmprint and obtain a better recognition effect.