A palmprint ROI extraction method for mobile devices in complex environment

Palmprint ROI (Region of Interest) extraction is an important procedure in palmprint recognition or authentication which mainly aims to obtain standardized images for feature extraction and matching. Locating contactless palmprint ROI in complex environment is a challenging work because of unstable external factors and relaxed operation environment. As to mobile palmprint recognition, the inherent problems mainly focus on hand postures, complex background, illumination variation and restricted processing capability. In this paper, a new palmprint ROI extraction (PRE) method is proposed which uses the modified convexity theory. PRE method abandons typical edge detection, centroid and coordinate procedure. Verification experiments on contactless palmprint databases demonstrate that PRE method is feasible and effective for mobile palmprint ROI extraction.