Active Shape Models과 타원을 이용한 손가락 끝 검출 및 추적

Detecting and tracking fingertips are significant for recognizing hand gestures. Generally, a lot of kinds of templates and skin color information are used to extract and trace feature points. However, it is difficult to find correct shapes and determine gestures. In this paper, we describe a method of detecting and tracking fingertips based on Active Shape Models(ASM) and a conic equation. ASM is good to obtain features of a specific object in images based on a trained model, and the ellipse equation helps a template follow the object without skin color information. We show how well fingertips are extracted and traced through some experiments.