Real-Time Tracking and Recognition of Human Hands from a Sign-Language Image Sequence

초록

This paper is about the way of hand-region detection and recognition which is most meaningful in sign-language. Gray scale image processing is required to avoid inconvenient sensor, for familiar with user, and for common use. For recognize the sign-language of deaf-mutes, preprocessing; detection of hand region in complex background images, is very important. With detection of hand region, it is possible that increase the recognition rate of hand region, and accelerate image processing. The proposed method is obtaining the image through subt ract one image from another sequential image, measuring the entropy, separating hand region from image, tracking the hand region. Then, the recognition process for movement pattern is conducted by simple HMM.