Analysis of Dorsal Palm Vein Pattern Recognition System

Individuals classification and recognition processes are a substantial growing field in many industry fields. In this paper, we presented a dorsal palm vein pattern recognition approach. Two approaches are presented. The first approach used the Principal Component Analysis (PCA) to extract features from images then the Multi-layer perceptron neural network for recognition step. The second approach Bag of features (BOF) used the Speeded-Up Robust Features (SURF) to extract local features from the training set for interest point selection then clustered in a representation set. The Support Vector Machine (SVM) technique is used in the classification phase. A comparison between the two approaches is proposed to observe the best approach that archived the higher classification accuracy. Here the dorsal palm vein images of the PUT database are used. The experiments show that the use of BOF is much better than PCA and MLP. Our experimental is able to recognize humans with accuracy 98% based on the BOF method.