Abstract: The small sample biometrics recognition is a difficult problem in real-world applications because the limited training samples can not lead to satisfactory recognition accuracy.So in this paper，a novel method is proposed by using Improved two-Directional Local Preserving Projections（I2DLPP） for feature extraction and Support Vector Machine（SVM） for classification，in small sample palmprint recognition.I2DLPP，an improved algorithm of 2DLPP by projecting 2DPCA and 2DLPP in the row and column directions simultaneously to reduce the computation complexity and the final feature dimensions；and the nearest-neighbor graph is constructed in which each node corresponds to a column in the image matrix.SVM is proven to be an effective tool for small sample biometrics recognition.The combination of I2DLPP and SVM can improve recognition performance significantly for small sample palmprint recognition.The experimental results demonstrate the effectiveness of the proposed method.

Key words: two-Directional Local Preserving Projections（2DLPP）, Support Vector Machine（SVM）, small samples, palmprint recognition

摘要： 小样本生物识别是现实应用中一个较难解决的问题，通过有限训练样本很难得到满意的识别结果。因此，提出了一种新的小样本掌纹识别方法，利用改进的二维局部保留映射（I2DLPP）提取特征，并用支持向量机（SVM）分类。改进的二维局部保留映射是通过同时在行和列方向上进行2DPCA和2DLPP的投影实现的，从而降低了计算复杂度与特征维数；并且构建最近邻图是以图像内部的列为节点，保留更多内部流形结构，改善了识别效果。SVM是针对小样本识别的非常有效的分类工具，将两者结合可以显著提高小样本掌纹识别精度。实验结果证明了该方法的有效性。

关键词: 二维局部保留映射, 支持向量机, 小样本, 掌纹识别