Abstract: In this paper, an algorithm for a fingerprint closed pore extraction is proposed. A closed pore is considered as a segment of binarized fingerprint image. Segment contains maximal information about a pore shape, orientation or other significant features. The proposed algorithm is based on the consecutive performance of some simple and well known image processing procedures, namely image binarization, segmentation, inversion, whitening etc. Segmentation is a process of splitting an image into non-overlapping partitions with connected pixels of the same intensity interval. After segmentation a pore is presented as a white segment in a black background. Inversion transforms the white pore segment into a black segment. Whitening is an operation to change pixels of the segment of certain size to pixels of intensity 255. This operation deletes black pores from the inverted image. Thus we can extract all the pores by comparing the intermediate images. The proposed algorithm consists of mentioned operations applied by appropriate choosing of thresholds. An example of application of described algorithm to show the effectiveness of our approach to the pore extraction problem is given.

Keywords: fingerprint, closed pores, segmentation, binarization, inversion.

ACM Classification Keywords: Image Processing and Computer Vision

Link:

SEGMENTATION BASED FINGERPRINT PORE EXTRACTION METHOD

David Asatryan, Grigor Sazhumyan

http://www.foibg.com/ijima/vol01/ijima01-2-p04.pdf