摘要
在对虚拟现实人物进行视景重构和三维重建中,为了提高虚拟人物图形生成的逼真性,需要结合图形与图像处理方法进行虚拟人物图形的三维跟踪渲染和重构,文章提出一种基于真实感约束的VR虚拟人物图形生成方法,采用图像边缘轮廓检测和纹理跟踪渲染方法进行虚拟人物图像采集和视景生成,并结合小波降噪方法对虚拟人物图形进行降噪滤波处理,提高人物图形成像的逼真度。在重构的像素特征空间中进行信息融合,为了提高VR成像的真实感,文章采用自适应纹理映射技术进行虚拟人物图形的跟踪渲染,将像素值加到三维实体模型上,生成具有真实感的VR虚拟人物图形。仿真结果表明,采用该方法进行VR虚拟人物图形生成的逼真度较好,输出图形的峰值信噪比较高,说明VR成像的质量较好,性能优于传统方法。
For the intention of improving the fidelity of the virtual characters and graphics generation in the virtual reality characters for scene reconstruction and 3D reconstruction, a combination of graphics and image processing method for virtual character graphics rendering 3D tracking and reconstruction is necessary. This paper proposed a method of VR virtual character graphics generation based on realistic constraints, in which will tracking image edge detection and texture of the virtual character image acquisition and visual simulation company with virtual character graphics noise filtering processing through wavelet denoising to improve the character graphics imaging fidelity. Information fusion had been used in pixel feature space reconstruction, the virtual character graphics using adaptive texture mapping technology tracking rendering, combined with the pixel value-added 3D model in order to enhance the authenticity of VR imaging and generate realistic VR virtual character graphics. Simulation result shows that this method works well in VR virtual character generation, and the peak signal noise of the output image is relatively high, which indicate that both the imaging quality and performance of VR and are better than traditional method.
作者
覃海宁
Qin Haining(Information Engineering Department of Guangxi Economic＆Trade Vocational Institute, Nanning 530021, China)
出处
《无线互联科技》
2017年第10期141-146,共6页
Wireless Internet Technology
基金
2015年度广西高校科学技术研究项目
项目名称:基于GPU的实时图形渲染引擎新技术研究
项目编号:KY2015YB446
关键词
真实感约束
虚拟现实
虚拟人物图形
成像
realistic constraints
virtual reality
virtual figure
imaging
