Uneven Background Extraction And Segmentation Of Good, Normal And Bad Quality Fingerprint Images

Abstract

In this paper, we have considered a problem of uneven background extraction and segmentation of good, normal and bad quality fingerprint images, though we propose an algorithm based on morphological transformations. Our result shows that the proposed algorithm can successfully extract the background of good, normal and bad quality images of fingerprint and well segment the foreground area. The algorithm has been tested and executed on FVC2002 database and the performance of proposed algorithm is evaluated through subjective and objective quality measures. This algorithm gives good and promising result and found suitable to remove superfluous information without affecting the structure of fingerprint image as well as reduces the storage space for the resultant image upto 77%. Our results will be useful for precise feature extraction in automatic fingerprint recognition system.