Fingertip Detection for Hand Pose Recognition

Abstract

In this paper, a novel algorithm is proposed for fingertip detection and finger type recognition. The algorithm is applied for locating fingertips in hand region extracted by Bayesian rule based skin color segmentation. Morphological operations are performed in the segmented hand region by observing key geometric features. A probabilistic modeling of the geometric features of finger movement has made the finger type recognition process significantly robust. Proposed method can be employed in a variety of applications like sign language recognition and human robot interactions.