Fingerprint & Palmprint Segmentation by Automatic Thresholding of Gabor Magnitude

Fingerprint & Palmprint are widely used biometric traits for human identification and authentication. Fingerprint & Palmprint Recognition Systems need feature images without background; this requires segmentation of input image. In this paper we have proposed a segmentation technique which is applicable for segmentation of both fingerprint and palmprint data. A unique feature of this method is that the threshold required for the segmentation is automatically calculated and need not to be determined empirically or manually. Separate threshold is generated dynamically for each input. Proposed method can be used in preprocessing stage of automatic fingerprint and palmprint recognition systems.