一种基于Leap Motion的非参数RDP检测算法  

A NON-PARAMETRIC RDP DETECTION ALGORITHM BASED ON LEAP MOTION

作　　者：张雪鉴 刘宏哲 黄先开[2] 袁家政 

机构地区：[1]北京市信息服务工程重点实验室,北京100101 [2]北京开放大学

出　　处：《计算机应用与软件》2017年第6期169-175,186,共8页Computer Applications and Software

基　　金：国家自然科学基金项目(61271369;61571045);国家科技支撑项目(2014BAK08B;2015BAH55F03);北京市自然科学基金项目(4152018;4152016);北京联合大学人才强校计划人才资助项目(BPHR2014A04)

摘　　要：随着计算机软硬件技术的进步和应用的普及,人机交互技术在博物馆领域中扮演着越来越重要的角色,并且受到了学术界和产业界的高度重视。尤其是Leap Motion体感控制器的出现,使人机交互的应用范围更加广泛与成熟,操作者可以通过非接触式的方式对设备进行操作,而无需使用触控屏、鼠标、键盘等外部设备,令人机交互方式更加友好、便捷。为了提高手势识别的准确性与实用性,提出一种基于Leap Motion的非参数RDP检测算法应用在手势识别中,并与Ramer-Douglas-Peucker(RDP)算法进行比较。实验证明使用非参数RDP检测算法可以有效地识别手势并且具有很好的自适应性。With the development of computer hardware and software and the popularization of application, human- computer interaction （HCI） is playing an increasingly important role in the field of museum, and has been attached great importance by academia and industry. In particular, the emergence of Leap Motion somatosensory controller, the application of human-computer interaction more extensive and mature. The operator can operate the device in a non-contact manner without using external devices such as a touch screen , a mouse, a keyboard, etc. , so that the human- computer interaction is more friendly and convenient. In order to improve the accuracy and practicability of gesture recognition, a non-parametric Ramer-Douglas-Peucker （RDP） detection algorithm based on Leap Motion is proposed and applied to gesture recognition and compared with RDP algorithm. Experiments show that the non-parametric RDP detection algorithm can effectively identify the gesture and has good adaptability.

关 键 词：非参数RDP检测算法 手势识别 Leap MOTION 人机交互 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]