中文摘要：

利用手指静脉造影识别人的身份已发展成为自动身份识别领域的一种重要方式,并广泛应用于多种实际场景。为了提升手指静脉图像的识别率,提出了一种基于随机森林的手指静脉识别方法。先将灰度化的手指静脉图像利用圆形等价模式LBP算子计算出编码值,然后通过子窗口直方图降维获取用于识别的特征向量,用随机森林集成分类器完成分类识别。通过实验对比分析,结果表明该方法能达到较理想的识别效果。

英文摘要：

It is an important method in the field of automatic identification using the finger vein recognition. It is widely used in many practical situations. In order to improve the recognition rate of finger vein image, the new finger vein recognition method based on random forest was proposed. Firstly, the codes of gray finger vein image were calculated by the circular Uniform Pattern LBP operator. Secondly, the feature vectors were extracted using the histograms of sub windows, then which were classified based on random forest ensemble classifier. The results show that the proposed method can achieve better recognition results than the experiments of references.