Fingerprint Image Segmentation Using Block-Based Statistics and Morphological Filtering

Fingerprint segmentation is meant to separate the foreground region of a fingerprint image from its background region. This paper presents a block-based segmentation scheme which is executed in two passes. In the first pass, two sets of regions of interest (ROI) are identified separately using (i) morphological open-close filters and (ii) a statistical measure namely coefficient of variation (CV). These sets of ROIs are combined together to identify the overall ROI. In the second pass, a block-wise region shrink–merge technique, which employs a sequential combination of parameters like CV and average gray value, is applied to construct the final segmented image. The proposed method has been implemented and tested on a set of real fingerprint images and the experimental results visually establish the effectiveness of the proposed method. Besides, a comparative study based on some quantitative measures is furnished to verify the accuracy of the proposed segmentation algorithm.