基于摄像头的近距手势识别系统的设计

Design of Embedded Camera-based Gesture Recognition System in Short Distance

随着电子技术的不断发展,人机交互方式也在不断得到转变,手势识别作为其中一项典型应用正吸引越来越多人的关注;为了进一步对该项技术进行研究和应用,在以STM32作为微处理器的嵌入式平台上,利用摄像头进行手势图像数据采集,通过FIFO和SRAM对数据进行缓冲和存储,对采集的图像进行了灰度转换、背景更新、差影分割、二值化、噪声去除等算法处理,完成了近距离范围内对运动手势的实时定位和基本识别;此基础上对俄罗斯方块游戏进行了手势控制,实现了手势识别技术在人机交互中的应用,体现出手势操作的便利性和全新用户体验的前景价值。

With the continuous development of electronic technology, the traditional human-computer interaction is also continually chan- ging. Gesture recognition as one of the most typical applications attracts more and more attention. For further research and application of gesture technology, the embedded platform with STM32 MCU is established. The camera is used for gesture acquisition, while the FIFO and SRAM are used for data buffering and storage. The gesture images are processed with corresponding algorithms, including gray scale trans- formation , background updating, image binarization and denoising processing etc. , and finally it achieves the real-time location and identifi- cation of gesture in short distance. The game of Russia box is controlled by gesture, and the effectiveness of recognition system is verified. It shows the convenience of gesture operation and the prospect value of new experience for users.