포토그래메트리 기반 3D 스캔 개선을 위한 적외선 카메라 활용

초록

In digital content creation, 3D models and spaces are utilized in various areas. Among them, photo-based 3D scanning technology, which uses a method of creating three-dimensional models based on 2D images, is cost-effective and versatile in many situations. However, it is challenging to obtain clean photos without noise in low-light environments using visible light cameras. This paper focuses on the research that shows infrared imaging using infrared lighting, rather than standard lighting in low-light environments, provides higher usability for 3D reconstruction in such conditions. The method proposed in this study uses artificial intelligence to improve the noise that is not completely eliminated in images. Subsequently, the created models are made usable in other 3D software through a 3D reconstruction program. These models can be applied and utilized in digital twin and other virtual environment contents.

키워드