修补
人工智能
计算机视觉
RGB颜色模型
计算机科学
深度图
渲染(计算机图形)
计算机图形学(图像)
图像(数学)
作者
Zhang Yan,Wang Jian,Che Dong-Juan
出处
期刊:Advances in intelligent systems and computing
日期:2020-01-01
标识
DOI:10.1007/978-981-13-9406-5_55
摘要
The rendering technology-based depth map can greatly reduce amount of data in encoding, decoding, and transmission. Similarly, this method is important for synthesizing virtual viewpoints. RGB and depth (RGB-D) image can be acquainted by Microsoft Kinect camera. But hole and noise problems for depth map must be inpainted and improved. In this paper, based on background of RGB-D image and frame difference, the depth map is inpainted. The experimental results show that our method can obtain the better depth map.
科研通智能强力驱动
Strongly Powered by AbleSci AI