Semi-automated luminance map re-projection via high dynamic range imaging and indoor space 3-D reconstruction

초록

Luminance-based metrics have been used in visual comfort studies that leverage high dynamic ranging imaging (HDRI) techniques. This paper presents a semi-automated robust framework for monitoring luminance within the FOV of occupants using a HDRI camera installed in a non-intrusive position. A new HDRI pipeline is presented, and a Structure-from-Motion and Multi-View-Stereo (SfM-MVS) photogrammetry framework is employed to automatically reconstruct the 3D room geometry and estimate occupant head-poses. The only required "commissioning" effort is to capture a short video of the room and take a few photos from the occupant position using a smartphone. Retrieved 3D context enables the back-projection of the camera-captured luminance distribution into 3D spaces, which is re-projected to occupant FOVs with the estimated head poses. The framework was implemented in a real daylit office and showed reasonable performance in re-projected indoor luminance map as perceived from the occupant position, except for sunlit projections from unshaded windows.

키워드