Chinese Static Sign Language Recognition Based On Hough Transform And Neural Networks

In recent years, with the development of society, accessibility of information on research and development is getting more and more noticed. Sign language recognition and sign language synthesis technology is one of the very important technologies. The development of the technology provides a good channel for the natural communication between the deaf and the normal.Sign language is the basic method of communication between the deaf. In order to make the deaf and the normal natural exchange, we need to translate sign language into natural language, and vice versa.Chinese sign language recognition based on Hough transform and neural networks system allows users to have a natural way of interaction. Firstly acquisition static gestures and pretreatment, that is standardization, gray, Canny edge detection etc; then use Hough transform to extract feature; finally use the BP neural networks trained to recognize static gestures. Hough transform has the good nature that is not sensitive to local defects, the robustness of random noise and suitable for parallel processing, real-time application etc, and without the impact of rotation, translation, expansion. So it is very meaningful for the research and application of sign language recognition by using Hough transform.This algorithm is to test gestures by using 20 samples, and collect 10 gesture pictures for every gesture. The experiments showed that the correct rate of Chinese static gesture recognition was 91.93 percent by using the algorithm.