Computer Games In The Gesture Recognition Technology Based On Computer Vision Research

In recent years, with the development and mature of computer technology, more and more researchers turn their attention to the improvement of the human-computer interaction and research into the human-computer interaction in accordance with interpersonal communication habits.Because the gestures as a communication tool has been widely used in people’s daily lives, the gesture recognition technology has been rapidly developed in this context. Gesture recognition technology based on computer vision has the advantage of man’s hand in nature state and gesture information collection and entry simple. It is an effective way of human-computer interaction as interpersonal communication habits. Men’s hand has significant individual differences. And external environment has a significant impact on gesture recognition technology. The region similar to skin color exists in complex background. These things cause a great deal of difficulty to gesture recognition technology based on computer vision. Research into this subject becomes a challenging and multi-disciplinary research. This article achieves the vision-based gesture recognition with the OpenCV development package under the Microsoft Visual Studio2008environment. The hand information and hand’s motion information extracted by gesture recognition as the parameter control the game programmed by OpenGL. Through the application of gesture recognition, we try to change the mode of human-computer interaction and make the mode of human-computer interaction more suitable for interpersonal communication habits.Through the camera to capture video images of people’s hand, we do preprocessing the image. We convert the images’color space and because of the skin color’s aggregation characteristics of the color in the color space, we can extract and cut up the gesture from the images. In extraction and segmentation of the gesture process, according to the color component value, we convert the images into gray value images and then set a reasonable threshold to convert grayscale images into binary images. Through expansion algorithm, corrosion algorithm opening operation and closing operation, we eliminate the interference point. We take the density distribution feature as the gestures characteristic to contrast with standard template to identify specific gestures. We take the hand information and hand’s motion information extracted by gesture recognition as the parameter to control the2D game programmed by OpenGL. This can replace the mouse and keyboard on the operation of the game.This paper aims to achieve the vision-based gesture recognition technology applied to computer games. People can control the game through gestures. The main value of the study is reflected in the following:In the gesture extraction and segmentation process, convert gesture image to HSV color space, split out the skin color using chrominance information of the pixel in this space. This method can exclude the interference of light, background and other environmental factors and individual differences factors and greatly enhance the efficiency of gesture extraction and segmentation accuracy. Taking density distribution feature as a gesture characteristic has the advantage of rotation, translation and scaling invariance. This method can weaken the impact of men’s hand size, offset from the center of the screen is different and the men’s hand distance from the camera on the accuracy of identification. Vision-based gesture recognition technology applied to computer games is the exploration of the change of human-computer interaction.