Abstract
In this paper, we propose a new method to solve sign language recognition problem using appearance-based features. Particularly, Local Binary Patterns (LBP) are employed to describe the texture and the shape of sign language images. The feature vector resulted from LBP operator is further reduced using Principal Component Analysis (PCA). The appearance-based features are classified using Hidden Markov Models (HMM). The performance of the proposed method is measured using Arabic Sign Language (ArSL) database. The proposed method does not rely on the use of data gloves or other means of input devices, and it allows the deaf signers to perform gestures without imposing any restriction on clothing or image background. Using LBP and PCA features, a recognition rate up to 99.97% was achieved for signer dependent recognition.