Automatic Segmentation of the Face and Hands in Sign Language Video Sequences

In this paper, we present an automatic hand and face segmentation algorithm based on color and motion information. The algorithm is proposed for the content-based representation of sign language video sequences, where the hands and face are perceptually significant. Our hand and face segmentation algorithm consists of three stages, namely skin-color segmentation, change detection, and video ob-ject plane generation. In skin-color segmentation, we model the skin color as a normal distribution and classify each pixel as skin or non-skin based on its Ma-halanobis distance. The aim of change detection is to localize moving objects in a video sequences. A statistical variance test is employed to detect object motion between two consecutive frames. Finally, the results from skin-color segmentation and change detection are analyzed to segment the face and hands. The perfor-mance of the algorithm is illustrated by simulations carried out on the silent and Irene test sequences.