Keypoint Selection Algorithm for Palmprint Recognition with SURF
Anca Ignat, Ioan Păvăloi · Procedia Computer Science · 2021
In this article we employ SURF features for palmprint characterization and recognition. We developed an original method (we named it FIKEN) for computing a fixed number of SURF keypoints. The classification process is based on the nearest neighbor ratio procedure and a distance measure that uses the locations of the matched keypoints. The experiments were performed on three well-known palmprint databases, IITD, CASIA and GPDS. The obtained experimental results show that the proposed FIKEN method gave very good recognition results that outperformed state-of-the-art results for these collections of palmprint images.