Efficient Local Representations for 3D Palmprint Recognition.pdf

1 Efficient Local Representations for 3D Palmprint Recognition Bing Yang1, Xiaohua Wang2 1School of Computer Science and Technology, Hangzhou Dianzi University, Xiasha Higher Education Zone, 310018, Hangzhou, China 2China Jiliang University, Xueyuan Street, Xiasha Higher Education Zone, Hangzhou, China Phone Number: +86115305712325 E-mail: ybily061821@126.com Abstract Palmprints have been broadly used for personal authentication because they are highly accurate and incur low cost. Most of the previous work have focused on two dimensional (2D) palmprint recognition in the past decade. Unfortunately, 2D palmprint recognition systems lose the shape information when capturing palmprint images. Moreover, such 2D palmprint images can be easily forged or affected by noise. Hence, three dimensional (3D) palmprint recognition has been regarded as a promising way to further improve the performance of palmprint recognition systems. In this paper, we have developed a simple, but efficient method for 3D palmprint recognition by using local features. We first utilizes shape index representation to describe the geometry of local regions in 3D palmprint data. Then, we extract local binary pattern (LBP) and Gabor wavelet features from the shape index image. The two types of complementary features are finally fused at a sore level for further improvements. Experimental results on the Hong Kong Polytechnic 3D palmprint database, which contains 8000 samples from 400 palms, illustrate the effectiveness of the proposed method. Keywords: 3D palmprint recognition, shape index, local representation, feature fusion. 1. Introduction Nowadays, information security has become a critic issue, and receives extensive attention. Most of conventional authentication methods, such as integrated-circuit cards, passwords, identity cards, etc., have various defects. For instance, passwords may be forgotten or compromised, identity cards may be lost or forged. Therefore,