Isolated dynamic Persian sign language recognition based on camshift algorithm and radon transform

Sign language is the initial tool for communication of deaf people in their everyday life. A lot of attention has recently been assigned to sign language recognition (SLR) by researchers in various domains such as computer vision, image processing and pattern recognition. Sign language gestures are divided in two groups, static and dynamic. The former includes the alphabets and the latter presents particular concepts. This paper presents a system for recognizing Persian sign language (PSL) in color video sequences. The system includes three main parts: tracking hand using continuously adaptive mean-shift (CAMSHIFT) algorithm, feature extraction using radon transform and discrete cosine transform (DCT). Finally to evaluate the impact of feature extraction technique on recognition rate, four different classifiers include minimum distance (MD), K-nearest neighbor (KNN), neural network (NN), and support vector machine (SVM) are used. The experimental results show that the suggested system is successfully able to recognize Persian gestures.