基于手势识别技术的交互式虚拟摄影系统设计与实现

摘要

传统虚拟摄影系统使用手动式控制识别,无法实现短距离交互式非触感识别,为此设计基于手势识别技术的交互式虚拟摄影系统.使用Kinect深度图像模型对手势图像进行分割预处理,完成手势关键节点定位;使用Kinect属性传感器剥离环境条件以及手势表面噪音数据;以数据编程形式对手势特征进行提取并标记关键节点;计算定位延迟,实现交互式虚拟手势姿态定位.试验结果表明,设计的交互式虚拟摄影系统比传统虚拟摄影系统的Miuyg参数高出20%,说明设计的系统具备极高的有效性.%The traditional virtual photography system uses the manual mode for recognition control,which can't realize the short-distance interactive non-tactile recognition. Therefore,an interactive virtual photography system based on gesture recogni-tion technology was designed. The Kinect depth image model is used to perform the segmentation and preprocessing for the ges-ture image to locate the gesture key nodes. The Kinect attribute sensor is used to eliminate the environmental conditions and ges-ture surface noise data. In the form of data programming,the gesture characteristics are extracted and key nodes are marked. The positioning delay is calculated to realize the interactive virtual gesture positioning.The experimental results show that the Miuyg parameter of the interactive virtual photography system is 20% higher than that of the traditional photography system,which indi-cates that the designed system has higher effectiveness.