基于OPENCV的手势识别系统的设计与实现

Design and Implementation of Gesture Recognition System Based on OPENCV

摘要:

人与计算机的交互技术是一种新型的计算机技术,且逐渐演变为一种主流技术和计算机领域的技术热点；为了能够更好地识别手势和跟踪手势的运动轨迹,提出了基于OPENCV的手势识别系统,系统引入了OPENCV计算机视觉库,OPENCV作为优秀的计算机视觉库,为设计的实现提供了便捷的代码,利用OPENCV技术中的图像处理算法,首现通过摄像头采集数据图像,并对采集到的图像进行一系列的缩放,去噪以及锐化等处理,然后对人体手势建立肤色模型,然后经过灰度阈值化来转换成二值图像,得到手轮廓的数据图像后,采用轮廓匹配方法识别出手型；最后通过10种基本的手势模型对比验证了本系统具有一定的实时性,并且识别率可以达到95%以上。

Abstract:

Interaction of human and computer is a computer model, and gradually evolved into a hot technology in a mainstream technology and computer science. In order to trajectory identification and tracking of gestures can be better, the gesture recognition system based on OPENCV is proposed, the system introduces the OPENCV computer vision library, computer vision library OPENCV as excellent, provides convenient code for the realization of the design, the use of image processing algorithms in OPENCV technology, is the first to establish the skin color model of human gesture then, through gray level thresholding to convert binary value image, then the morphological transformation, finally obtains the data image hand contour, the contour matching method, the final recognition hand type, and can effectively analyze the motion trajectory. The experiments verify the effectiveness and practicability of the system.