Data associated with the publication: C3VDv2: 3D colonoscopy video dataset with enhanced realism

3D computer vision techniques could improve the diagnostic value of colonoscopy, but the lack of 3D colonoscopy datasets for training and validation hinders their development. This paper introduces C3VDv2, a high-definition Colonoscopy 3D Video Dataset with enhanced realism designed to facilitate the quantitative evaluation of 3D colon reconstruction algorithms. 192 video sequences were captured by imaging high-fidelity phantoms from two colon anatomies. Ground truth depth, surface normals, optical flow, occlusion, six-degree-of-freedom pose, coverage maps, and 3D models are provided for 169 colonoscopy videos. Eight simulated screening colonoscopy videos acquired by a gastroenterologist are provided with ground truth poses. The dataset includes 15 videos featuring colon deformations for qualitative assessment. C3VDv2 elevates realism by incorporating diverse and challenging scenarios for 3D reconstruction algorithms commonly encountered during colonoscopy procedures, such as the presence of fecal debris, mucous pools, blood, debris obscuring the colonoscope lens, en-face views, and fast camera motion. The comprehensive nature of C3VDv2 offers a robust and diverse dataset for the quantitative and qualitative evaluation of 3D reconstruction algorithms. . This work was supported in part with funding and products provided by Olympus Corporation of the Americas. Although the agreement states a Sponsored Research Agreement, Olympus is funding, but not sponsoring this research. (2024-06)