Abstract：In order to suppress the palm print image noise effectively and extract the palm print features accurately, a novel de-noising method based on 2-D EMD and wavelet thresholding was proposed. Firstly, the preprocessed palm print image containing noise was decomposed into some IMFs by 2-D EMD, and then the first several IMFs were processed by using wavelet thresholding de-noising method; finally, the image was reconstructed through adding the processed IMFs and the residual component. Experimental results show that the proposed method has a good effect on suppressing noise. Compared to the wavelet thresholding and 2-D EMD de-noising, the method not only has advantages of more sufficiently retaining edge and detail information while de-noising, but also achieves superior PSNR (peak signal-noise-ratio) of the reconstructed image.
［1］吴介, 裘正定. 掌纹识别中的特征提取算法综述［J］. 北京电子科技学院学报, 2005, 13(2): 86－92.
［2］ Xu Guan-lei,Wang Xiao-tong,Xu Xiao-gang. Neighborhood limited empirical mode decomposition and application in image processing［C］// IEEE Proceedings of the 4th International Conference on Image and Graphics, Washington,DC, USA, 2007, 149－154.
［3］ 练秋生, 刘春亮. 基于Gabor滤波器和LBP 的分级掌纹识别［J］. 计算机工程与应用, 2007, 43(6): 212－215.
［4］ He Lu-lu, Wang Hong-yuan. Spatial-variant image filtering based on bidimensional empirical mode decomposition ［C］//IEEE Proceedings of the 18th International Conference on Pattern Recognition, Washington,DC, USA, 2006, 2, 1196－1199.
［5］ Nunes J C, Guyot S, Del′ echelle E. Texture analysis based on local analysis of the Bidimensional Empirical Mode Decomposition ［J］. Machine Vision and Applications, 2005, 16:177－188 .
［6］ Damerval C, Meignen S,Perrier V. A fast algorithm for bidimensional EMD ［J］. IEEE Signal Processing Letters, 2005, 12(10): 701－704.
［7］ 戴桂平, 刘彬. 基于白化检验和3σ准则的小波阈值去噪算法［J］. 计量学报, 2007, 28(2):158－161.
［8］ 陈智, 黄玲玲. 基于ICA和BP神经网络相结合的掌纹识别 ［J］. 北京航空航天大学学报, 2008, 34(3): 291－293.
2011, Vol. 32 
