The Research Of Sign Language Recognition Technology Based On Complex Background

As the society develops,information accessibility and relevant research have become more and more important.Sign language recognition and synthesis is one of the most typical research and technology in this area.It helps to establish a smooth channel between deaf person and normal ones in communication.As a communication tool,sign language has been used by deaf people in their conversation and contact.In order to find an easier and better way to make the communication between deaf group and normal people,sign language needs to be translated into natural language that normal people could comprehend.The research and implementation of sign language recognition has significant academic value and broad application prospect.In this paper,as to the hand gesture segmentation,we use the commonly used self-adaptive OTSU algorithm;as to the moving object detection,we proposes movement target region examination method which combined the background-difference method with double-difference method,it can be used to estimate contour detection and moving track detection;as to the feature extraction,using the Movement Non-political reform method to calculate centre of moving gesture,combined with moving target examination,we conclude the track of moving gesture,which would be used to parametric training and recognition.Research results show that the proposed methods can work well for the recognition of sign language.Meanwhile we find from the research that besides the reason of algorithm,the quality of camera has a remarkable impact on the recognition result,and it would direct the following research.