Panoramic Ceiling Image Synthesis Method Prioritizing Fixture Outlines using an Omnidirectional Camera

Daiki Kohama, Yoshiteru Nagata, Kazushige Yasutake,Kenta Urano,Shin Katayama, Takuro Yonezawa,Nobuo Kawaguchi

2023 FOURTEENTH INTERNATIONAL CONFERENCE ON MOBILE COMPUTING AND UBIQUITOUS NETWORK, ICMU(2023)

引用 0|浏览0
暂无评分
摘要
The manual creation of a ceiling plan consumes many human resources to confirm the current state of existing buildings for the renovation. Especially the positions and the types of existing fixtures are important. In this paper, we propose a synthesis method for a panoramic ceiling image using a video shot by an omnidirectional camera. We utilize a Visual SLAM method to map the positions of omnidirectional video frames including the ceiling. We prioritize the clear depiction of fixture outlines for easier identification of the types of ceiling fixtures. To draw clear fixture outlines in the resulting image, each fixture is referred from the single video frame that captures it most clearly. In the experiment, we collected an omnidirectional video in the indoor area of approximately 273 square meters with multiple rooms. We evaluated the synthesized image for positional accuracy of the ceiling fixtures, and our proposed method demonstrated higher accuracy compared to the baseline.
更多
查看译文
关键词
Omnidirectional Camera,Visual SLAM,Image,Synthesis,Ceiling Plan
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要