Texture Feature based Fingerprint Recognition for Low Quality Images

Fingerprint-based identification is one of the most well-known and publicized biometrics for personal identification. Extracting features out of poor quality prints is the most challenging problem faced in this area. In this paper, the texture feature based approach for fingerprint recognition using Discrete Wavelet Transform (DWT) is developed to identify the low quality fingerprint from inked-printed images on paper. The fingerprint image from paper is very poor quality image and sometimes it is complex with fabric background. Firstly, a center point area of the fingerprint is detected and keeping the Core Point as center point, the image of size w x w is cropped. Gabor filtering is applied for fingerprint enhancement over the orientation image. Finally, the texture features are extracted by analyzing the fingerprint with Discrete Wavelet Transform (DWT) and Euclidean distance metric is used as similarity measure. The accuracy is improved up to 98.98%