Manually selecting viewpoints or using commonly available flight planners like circular path for large-scale 3D reconstruction using drones often results in incomplete 3D models. Recent works have relied on hand-engineered heuristics such as information gain to select the Next-Best Views. In this work, we present a learning-based algorithm called Scan-RL to learn a Next-Best View (NBV) Policy. To train and evaluate the agent, we created Houses3K, a dataset of 3D house models. Our experiments show that using Scan-RL, the agent can scan houses with fewer number of steps and a shorter distance compared to our baseline circular path. Experimental results also demonstrate that a single NBV policy can be used to scan multiple houses including those that were not seen during training. The link to Scan-RL is available at https://github.com/darylperalta/ScanRL and Houses3K dataset can be found at https://github.com/darylperalta/Houses3K.
@article{arxiv.2008.12664,
title = {Next-Best View Policy for 3D Reconstruction},
author = {Daryl Peralta and Joel Casimiro and Aldrin Michael Nilles and Justine Aletta Aguilar and Rowel Atienza and Rhandley Cajote},
journal= {arXiv preprint arXiv:2008.12664},
year = {2020}
}
Comments
To be published in ECCV 2020 Workshops; typos in abstract corrected