基于局部纹理特征的三维掌纹识别研究

3D Palmprint Recognition Based on Local Texture Feature Sets

近年来,自动掌纹识别方法的研究吸引了越来越多的关注,已有的工作主要集中于二维掌纹识别。然而,二维掌纹图像存在着易伪造、抗噪能力差的缺陷,实际应用中会带来潜在的安全隐患。因此,三维掌纹识别被视为一种可行的解决方案来进一步提高识别的性能。基于局部纹理特征,本文提出一种有效的三维掌纹识别方法。该方法首先利用形状指数来描述三维掌纹的局部几何特征,接着提取形状指数图像的局部三值模式以及Gabor小波特征,最后在匹配分数层次上对这两种互补的局部纹理特征进行融合,随后的实验证明了融合特征较单独特征要好。在香港理工三维掌纹数据库上的实验结果表明,本文方法在识别率上要优于目前流行的其它三维掌纹识别方法,从而验证了本文方法的有效性。

Recent years have witnessed a growing interest in developing automatic palmprint recognition methods. Most of the previous works have focused on two dimensional （2D） palmprint recognition in the past decade. However, 2D palmprint images could be easily forged or affected by noise, causing potential security risks for practical applications. Therefore, 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 proposed an efficient 3D palmprint recognition method by using local texture feature sets. We first employ shape index representation to demonstrate the geometry characteristics of local regions in 3D palmprint data. Then, we incorporate rich local texture cues from two complementary sources-local ternary pattern （LTP） and Gabor wavelet to extract features from the shape index image-proving that the combination is more accurate than either feature set alone, and finally fuse them at a matching score level. Further experiments on Hong Kong Polytechnic University 3D palmprint database validate that our method outperforms existing state-of-the-art methods in terms of recognition accuracy, showing the effectiveness of our method.