Method of multi-modal biometric recognition using hand-shape and palmprint

A method of multi-modal biometric recognition, comprising the steps of:

• capturing a hand image of a user by using a digital camera;
• transforming the captured hand image into a digitized image by the digital camera;
• binarizing said digitized image;
• obtaining an orientational angle θ and location information of said binarized image;
• vertically aligning said binarized image, based on the orientational angle θ and the location information from binarized shape moments, using a rotation matrix;
• establishing a geometrical center of a palmprint image of said vertically aligned image based on selected pixel(s) with highest magnitude of distance transform;
• segmenting said palmprint image;
• extracting palmprint feature vectors of said palmprint image using discrete cosine transform;
• obtaining a hand shape image from said vertically aligned image, and extracting hand shape feature vectors from localized information of the hand shape image;
• selecting a subset of the hand shape feature vectors and a subset of the palmprint feature vectors;
• fusing the subset of said hand shape feature vectors and the subset of said palmprint feature vectors to form a fused feature vector;
• and
• classifying said fused feature vector.