Hand geometry and palmprint-based authentication using image processing

Abstract

Nowadays, the importance of biometric authentication has increased, since biometric identifiers are unique to every individual, and are more reliable than knowledge or ownership factors. Palmprint and hand geometry features are suitable for high-confidence personal authentication. Contactless authentication is another topic which has been researched for capturing a palm image in more sanitary conditions. The combined use of multiple biometric identifiers results in a more reliable authentication system. This paper proposes a method for contactless hand geometry and palmprint-based authentication. The proposed method uses simple image processing algorithms and geometric relations for keypoint detection and palm image extraction. Morphological operations, averaging filters and different methods for noise filtering are used for palmprint principal line extraction. The corresponding points of palmprint images are detected by template-based matching which is followed by further filtering of matching points. With the use of hand geometry features and palmprint matching mechanism, the collected information of the hand and palm is suitable for authentication.