Feature Extraction and Matching

Summary

This chapter revealed a line based feature extraction and matching strategy from the offline palmprint images. Obviously, line feature includes curves and straight lines. Based on our observation, the curvature of the principal lines is small enough to be represented by several straight line segments. Straight lines are easy to represent and match. The matching of palm lines can be performed by measuring the Euclidean distances between the endpoints of two line segments and the angles between these line segments. Experimental results show that this method is effective.

In addition, a classification method is proposed based on the singular points, such as loop, whorl and arch formed by ridges, found on the outside region of a palm. There are six typical cases of singular points in the outside region for palmprint classification. Experimental results show that around 97% correct classification can be achieved in a database of 354 offline palmprint images.