摘要
在众多影响3D舒适度的因素中,制作人员更多的会选择视差信息来检测、调整、优化3D节目的质量,因此3D电视素材的左右图像的视差分析成了关键问题。本文设计了一种基于双目立体视觉的3D电视视差检测系统,该系统应用了双边滤波及SGBM立体匹配算法,计算左右图像的视差,并根据视差与立体视觉舒适度的关系,将可能会引起观众不适的图像部分进行分级标注,便于后续的3D调整工作。
Production staffs prefer to choose horizontal disparity to detect, adjust, and optimize 3D program quality. Therefore disparity analysis becomes a key issue. Bilateral filtering and stereo matching algorithm SGBM are used to calculate horizontal disparity. Some parts of the image which may cause the audiences uncomfortable are marked according to the relationship between horizontal disparity and 3D visual com- fort. And this will bring convenience for subsequent adjustment.
出处
《中国传媒大学学报(自然科学版)》
2014年第4期24-28,共5页
Journal of Communication University of China:Science and Technology
关键词
立体视差
视差图
双边滤波
SGBM
stereoscopic disparity
disparity map
bilateral filtering
SGBM