Poor Quality Fingerprint Recognition Based on Wave Atom Transform

Abstract

Fingerprint is considered the most practical biometrics due to some specific features which make them widely accepted. Reliable feature extraction from poor quality fingerprint images is still the most challenging problem in fingerprint recognition system. Extracting features from poor fingerprint images is not an easy task. Recently, Multi-resolution transforms techniques have been widely used as a feature extractor in the field of biometric recognition. In this paper we develop a complete and an efficient fingerprint recognition system that can deal with poor quality fingerprint images. Identification of poor quality fingerprint images needs reliable preprocessing stage, in which an image alignment, segmentation, and enhancement processes are performed. We improve a popular enhancement technique by replacing the segmentation algorithm with another new one. We use Waveatom transforms in extracting distinctive features from the enhanced fingerprint images. The selected features are matched throw K-Nearest neighbor classifier techniques. We test our methodology in 114 subjects selected from a very challenges database; CASIA; and we achieve a high recognition rate of about 99.5%.