摘要
体感交互应用是通过新型的体感交互设备来进行人机交互操作,在虚拟仿真、项目展示等方面有着重要应用。为解决体感交互过程中由于骨骼抖动导致行为误判的问题,提出了在Kinect与UnrealEngine4成功对接的基础上,加入动态灰度预测算法处理数据,同时利用灵敏度分析对数据量进行优化,在不影响正确率的情况下减少响应时间,提高识别速率。经过实验,该方法相比于移动平均法等过滤器在正确率和效率方面有了较大的提升。
Somatosensory interactive application can realize human-computer interaction through new type somatosensory interactive equipment, which has important applications in virtual simulation, project display and etc. To solve behavior misjudgment problem caused by bone jitter during somatosensory interaction process, the paper puts forward, adding process data of dynamic gray-scale prediction algorithm on the basis of successful docking of Kinect and Unreal Engine 4, at the same time, optimizing data amount with sensitivity analysis, reduce response time and improve recognition rate without affecting accuracy. Experiment showes, the method has better improvement of accuracy and efficiency than filters of moving average method.
作者
王海学
杨淑婷
张阳
楼容
刘峰
WANG Hai-xue;YANG Shu-ting;ZHANG Yang;LOU Rong;LIU Feng(School of Education Science and Technology,Nanjing University of Posts and Telecommunications,Nanjing,Jiangsu 210023;Jiangsu Image Processing and Communication Key Laboratory,Nanjing,Jiangsu 210003)
出处
《软件》
2020年第3期110-113,218,共5页
Software
基金
江苏省大学生创新训练计划项目(201810293038Z)。
关键词
体感交互
灰度预测
骨骼抖动
Somatosensory interactive
Gray level prediction
Bone jitter
