基于改进凸分解的手势识别  

Hand Gesture Recognition Based on an Improved Convex Shape Decomposition

作　　者：潘志庚[1] 王顺婷 姚争为[1] PAN Zhi-geng;WANG Shun-ting;YAO Zheng-wei(DMI Research Center, Hangzhou Normal University, Hangzhou, Zhejiang 311121, China)

机构地区：[1]杭州师范大学数字媒体与人机交互研究中心,浙江杭州311121

出　　处：《北京理工大学学报》2018年第3期279-285,共7页Transactions of Beijing Institute of Technology

基　　金：国家自然科学基金重点资助项目(61332017);国家科技支撑计划项目(2015BAK04B05);国家重点研发计划资助项目(2017YFB1002803)

摘　　要：提出了一种新的基于Kinect的实时静态手势识别方法,主要贡献包括:提出了一种简易可行的、结合图像深度信息与肤色信息的手势区域检测与分割方法;提出了一种改进的凸分解算法,对手势区域进行近似凸形状分解,以得到表征手势特征的骨架信息;采用基于路径相似性的骨架图匹配算法对手势进行匹配以实现识别.针对特定手势集进行了对比实验,实验结果表明,本文方法在识别结果的准确率以及算法的效率上都有着良好的表现.Hand gesture recognition has become one of the most natural way of human communication with computer, However,the recognition accuracy and efficiency of the traditional methods still need to be improved.In this paper,a novel method was proposed based on Kinect for static gesture recognition in real time.Firstly,a simple and feasible hand detection and segmentation method was put forward based on depth information and skin color information.Then,a improved convex shape decomposition method was developed to obtain hand skeleton,and a gesture recognition method was put forward based on comparing the geodesic paths between skeleton endpoints.Finally,a comparative experiment was carried out for a specific gesture set.The experimental results show that,this method can provided a good performance in recognition accuracy and efficiency of the algorithm.

关 键 词：手势识别 凸分解 骨架图匹配 深度信息 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]