A Self-reliance Smoothness of the Vision-Based Recognition of Hand Gestures

The hand gestures are a natural and intuitive mode for human-computer interaction. The vision-based recognition of hand gestures is a necessary method for future human-computer interaction. On the other hand, the edge detection in the course of the hand recognition is also a key technique. In this paper based on the classical self-reliance smoothness operator, we propose a new method of midpoint threshold (the averaged grads method). According to this method, the optimal threshold and edge image of hand gestures can be got by using the first or each information of grads of reference image.