Chinese sign language recognition based on gray-level co-occurrence matrix and other multi-features fusion

We propose a novel approach for solving the Chinese manual alphabet in vision. Rather than focusing on local features and their consistencies in the images data, our approach aims at extracting both the global and local features of an image. Features calculated from gray-level co-occurrence matrix and other multi-features are introduced for the classifier to characterize the various visual properties of the images. Experimentation with 30 groups of the Chinese manual alphabet images is conducted and the results prove that these global and local visual features, such as correlation, entropy, etc. are simple, efficient, and effective for characterize hand gestures, and the SVMs method shows excellent classification and generalization ability in solving learning problem with small training set of sample in sign language recognition. We choose the linear kernel function for the SVMs and found the results to be very encouraging: the average recognition rate of 93.094% is achieved.