Palmprint verification and identification using pyramidal HOG feature and fast tree based matching

In this paper, a palmprint identification and verification approach based on Pyramidal Histograms of Oriented Gradients (PHOG) and fast tree based matching is presented. In the feature extraction stage, proposed local histograms of oriented gradient are extracted in each level or scale of the Gaussian pyramid of the palmprint. This matter helps to extract high contrast and reliable lines. In the identification stage, features of the query palmprint are extracted and identified by matching against the training database palmprints or templates with tree based matching. The tree based matching makes verification and identification fast. The proposed method is evaluated by experiments on Hong Kong PolyU palmprint database. The experimental results show that the proposed method achieves a high performance according to identification and verification rate, over 99.85% accuracy, and high identification speed, which takes less than 2 seconds.