基于Kinect的虚拟3D视频会议场景融合研究  

Implementation of Virtual 3D Video Conference Based on Kinect

作　　者：程宏伟[1] 蒋乐天[1] 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海200240

出　　处：《计算机与数字工程》2013年第2期212-213,247,共3页Computer & Digital Engineering

摘　　要：为了应用3D摄像头构建新型视频会议模式,将与会人物的3D信息提取出来,融合在虚拟会议场景中,来实现完整的3D会议效果。由Direct3D工具设计出固有的会议场景,采用微软发布的Kinect摄像头提取与会者的3D模型信息,发送到服务器端,完成融合后再返回终端显示。融合结果有较强现场感,人物模型真实准确。验证了系统的可行性和实用性。To build a new type of video conference mode with 3D camera,the 3D model of participator is distracted and fused into virtual scene.The integrated system consists of the conference room,designed through Direct3D tool,and Kinect 3D camera proposed by Microsoft.3D model from Kinect,is sent to the server and fused into conference environment.And then sent back to terminal PC display.The result feels real and accurate,and the system can be feasible.

关 键 词：KINECT 3D人物模型 实时渲染 场景融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]