Gesture Recognition Technology Research And Application Based On Computer Vision

With the development of computer technology and Internet, the human-computer interaction is becoming more and more important in people’s daily life. As image processing, pattern recognition, computer vision and related technology development, people began to study more intelligent, more natural, more in line with the human communication habits of human-computer interaction technology. As a natural, intuitive, easy to learn interactively, gestures plays an important role in communication. Therefore, the gesture recognition technology has become a research hotspot in the field of human-computer interaction.With the rapid development of related technology, gesture recognition based on computer vision technology has made great progress. Because of the complexity of the realistic environment and gestures variability, however, the existing gesture recognition technology there are still some defects and shortcomings.In view of the existing problems, in this thesis, the specific research work includes the following aspects:1.Hand detection segmentation: Through research analysis of characteristics of skin color detection method, template matching method, and based on the movement of the prospect of extraction technology in the application of hand detection segmentation, summarizes and puts forward a new algorithm. Through different methods combined effectively, realize the target under the condition of complex hand area accurately and efficiently.2. Static gesture recognition: In this thesis, considering the overall and the local two dimensions to the feature extraction of the gesture: The low dimensional four Hu moment feature is used to describe the gesture contour shape characteristics, and LBP features is used to describe gesture texture feature. At the same time, in view of the recognition algorithm based on Hu moment, K neighboring classification strategy adopted to carry out the improvement and enhancement algorithm robustness. On the basis of the above, this thesis puts forward a kind of gesture recognition algorithm based on feature fusion, and the experiment proves the algorithm is effective.3. Finally, this thesis completed the gesture recognition technology combined with actual application scenarios, implements a simple image browsing system based on gesture recognition. By using this system, the user through a few simple gestures can achieve the basic operation of pictures.