A Palmprint Recognition Algorithm Based on GIDBC

• Published in: Biometric Recognition pp. 258 - 265
• Main Authors: Jia, Mengqi, Wei, Weibo, Hong, Danfeng, Wang, Gang
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Fractal dimension; Fractal operator; Gabor transform; Improved differential box dimension; Palmprint recognition
• ISBN: 3319254162; 9783319254166
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-25417-3_31

As fractal dimension could not describe the information of palmprint accurately as a characterization, differential box dimension (DBC) is improved by using a custom fractal operator, and a novel palmprint recognition algorithm based on Gabor transform and improved differential box dimension (GIDBC) is proposed in this paper. Firstly, Gabor transform is used for palmprint images in frequency domain to get the information of multi-scale and multi-direction, and the ideas of block is used to divide palmprint images into blocks. Then every block’s feature vector is extracted by using improved differential box dimension (IDBC) algorithm, and features of all blocks are fused in the parallel. Finally, chi-square distance is used for classification. Compared with those traditional algorithms by experiments in PolyU palmprint database, recognition rate can reach 99.78%, feature extraction and matching time is 338ms, which demonstrates the validity and efficiency of the proposed algorithm.