Computer assisted segmentation of palmprint images for biometric research

One of the becoming popular biometric modalities is the palmprint. This biometric modality is rich with information, such as minutiae, ridges, wrinkles, and creases. This research team is interested to investigate the creases for biometric identification. The palmprint images in this research have been captured by using a commercially available consumer scanner. For each palmprint image, two square regions on the palmprint image are extracted for biometric identification purpose. One of the regions is from the hyperthenar region, while the another is from the interdigital region. Due to misalignment of the hand, the process of extraction of these regions is tedious and time-consuming. Therefore, in this paper, a computer-aided method has been proposed to simplify the extraction process. The user only needs to mark two points on the palmprint image. Based on these points, the palmprint image will be aligned, and those two regions are extracted automatically.