Detecting and tracking fingertips are significant techniques for recognizing hand gestures. Many patterns and skin color information are used to extract and trace features [1, 2]. However, to find correct shapes is difficult and there is limitation to express the diversity of models. In this paper, we describe a method of detecting and tracking fingertips based on Active Shape Models (ASMs) and an ellipse equation. ASM is an effective tool to obtain features of a specific object in images by using a trained model. 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.