Point and local neighborhood operations fusion for efficient fingerprint enhancement

Abstract

Fingerprint is an important evidence in criminal investigation process. It is considered as the most reliable biometric measure in forensic sciences. Many studies have been proposed in the subject of fingerprint image enhancement but most of them either worked on point based operations or local neighborhood operations. This paper presents an efficient fingerprint enhancement approach by combining the point and local operations. Point based adaptive histogram equalization method along with local neighborhood based median filter method improves the overall quality of fingerprint by removing noise from image and preserves its information content in better way than individual point or local operations. The image entropy is used as an objective assessment method for the analysis of quality of fingerprint image.