Forest, helmet-held 360 camera
This model is best viewed by selecting the first person mode in navigation setting and slow speed (eg 0 or 1) and navigating at the neighborhood of the series of gray squares. It is computed from videos taken by a helmet-held DIY multi-camera (four Gopro Hero 3 at 100Hz forming a 360 camera) and by walking during 4 minutes in a wood. The series of gray squares indicates the camera poses of 878 keyframes in the videos. See also one of the keyframes by selecting annotations.
The main steps of the reconstruction method are in my publications: synchronization and self-calibration (3DV‘15, CVIU‘17), structure-from-motion (BMVC‘07, IVC‘09), surface reconstruction (CVIU‘13 improved by ICPR‘14), curves are matched (PAMI‘02) and integrated in the surface.
The original textures are divided by 2 and stored in a 8k texture image. The multi-camera model is non-central, global shutter, with frame-accurate synchronization.