计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (24): 202-206.DOI: 10.3778/j.issn.1002-8331.1607-0031

• 图形图像处理 • 上一篇    下一篇

基于Leap Motion和Unity3D的虚拟沙画手势识别及交互

宁亚楠,杨晓文,韩  燮   

  1. 中北大学 计算机与控制工程学院,太原 030051
  • 出版日期:2017-12-15 发布日期:2018-01-09

Gesture recognition of virtual sand painting interaction based on Leap Motion and Unity3D

NING Yanan, YANG Xiaowen, HAN Xie   

  1. School of Computer Science and Control Engineering, North University of China, Taiyuan 030051, China
  • Online:2017-12-15 Published:2018-01-09

摘要: 虚拟仿真技术的快速发展及体感设备的不断更新为沙画动画这一全新的艺术创作形式带来新的灵感。针对沙画现场作画工序复杂的问题,结合Leap Motion设备和Unity3D开发环境完成手势识别并驱动虚拟手实现虚拟沙画效果。首先,依据Leap Motion捕捉到的手势坐标及方向信息提取手部关键点;然后提出角域划分的方法并引入新的特征向量,将其与提取信息串联作为手势分类依据;最后,根据自行定义的沙画手势语义驱动虚拟手完成虚拟沙画创作。实验证明,利用Leap Motion完成近距离手势识别效果较其他方法结果更加精准,实时性较高,手势跟踪稳定,虚拟沙画绘画过程沉浸感强。

关键词: 虚拟沙画, 角域划分, 手势识别, Leap Motion, Unity3D

Abstract: Rapid development of virtual simulation technology and constantly updated somatosensory equipments bring new inspiration for sand painting animation which is a creative art form. Because of the complex process of live sand painting, this paper combines Leap Motion devices and Unity3D development environments to accomplish hand gesture recognition, and then successfully performs live sand painting in virtual environments. Firstly, hand key points of coordinate and direction are extracted according to the gesture information captured by Leap Motion. Then the method of angle domain segmentation is proposed and a new set of feature vector is added, which is in series with fetched information as the basis of hand classification. Finally, defined gestures are utilized to drive virtual hand to complete sand painting. The experiments show that gesture recognition by Leap Motion in close range performs more accurate than other methods, gesture tracking is stable with high real-time performance and the virtual sand painting system has a much stronger sense of immersion.

Key words: virtual sand art, angle domain segmentation, posture recognition, Leap Motion, Unity3D