基于图像转换的全景平滑漫游技术

摘要

Panorama-based virtual reality systems have a key problem is that how to achieve the smooth transition of two adj acent view spaces.Aiming at this issue,the present situation of current panoramic roaming system was analyzed.Using the computer vision technology,a panoramic smooth roaming technology based on image conversion was proposed.First of all,photos were collected on the transition path.Then according to the camera imaging model,the correlation between space points and the pro-j ection points was established and the relationship between the images in the translational motion was determined.Finally,ima-ges with the non-sampling points were synthesized.Experimental results show that the method obtains relatively smooth transi-tion effects.%基于全景图的虚拟现实系统存在实现两个相邻视点空间的平滑过渡的问题，为此对目前全景漫游系统的现状及功能进行分析，利用计算机视觉技术，提出基于图像转换的全景平滑漫游技术。在过渡路径上采集照片，根据摄像机的成像模型，建立空间点和投影点的关联，确定图像在平移运动下的关系，合成非采样点的图像。实验结果表明，该方法获得了比较平滑的过渡效果。