Enhanced Palmprint Recognition via Curvi-Linear Anisotropic Gaussian Filter-Based Combined Differential Concavity and Infirmity Codes
Pawan Dubey, Tirupathiraju Kanumuri, Ritesh Vyas, Keerty Venkata Sri Ramachandra Murthy, Chandan Kumar Choubey, Durgesh Nandan · Traitement du signal · 2023
The inherent curvature in palm lines can pose challenges for palmprint recognition, particularly at lower resolutions where wrinkles become indistinguishable, leading to performance degradation.To address these issues, this study introduces a novel methodology employing curvi-linear anisotropic Gaussian filter-based Combined Differential Concavity and Infirmity (CDCI) codes.The use of curved filters has been proposed to represent curved palm lines more accurately, while anisotropic filtering is expected to enhance the extraction of blurred palm lines.The new representation, grounded in curvi-linear anisotropic Gaussian filtering, is posited to improve the recognition system's performance by effectively addressing these challenges.The proposed approach's effectiveness has been tested using the touchless IITD database and the contact-based PolyU 2D database.The experimental results suggest that the proposed methodology surpasses the performance of state-of-the-art coding-based procedures in palmprint recognition with the improvement of 3.82% and 36.36% in recognition rate and equal error rate.