Enhanced Fingerprint Image Compression using Deep Deterministic Policy Gradient

This paper presents a novel approach to fingerprint compression using the Deep Deterministic Policy Gradient (DDPG) algorithm. The goal is to achieve high compression ratios while maintaining image quality. The DDPG algorithm combines deep neural networks with reinforcement learning and is well-suited for complex control tasks. Experimental evaluations are conducted on a benchmark fingerprint dataset, comparing the proposed method with state-of-the-art compression techniques. Results show that the DDPG-based compression approach achieves competitive compression ratios while preserving important fingerprint details. The algorithm demonstrates robustness across diverse fingerprint samples. Computational efficiency is also evaluated, considering training and compression times. The findings indicate a balanced trade-off between complexity and performance, making the DDPG-based approach suitable for practical fingerprint compression applications.