RidgeSfM: Structure from Motion via Robust Pairwise Matching Under Depth Uncertainty
This work provides a more scalable and accurate structure-from-motion solution for large-scale indoor scene reconstruction, benefiting applications requiring dense 3D models from many images.
This paper addresses the problem of simultaneously estimating dense depth maps and camera poses for indoor scenes from a large set of images. The authors developed RidgeSfM, which directly outputs poses and dense reconstructions from an altered bundle adjuster, allowing it to align hundreds of frames, significantly more than recent deep alternatives limited to around 10 frames. RidgeSfM achieves superior performance compared to state-of-the-art large-scale SfM pipelines.
We consider the problem of simultaneously estimating a dense depth map and camera pose for a large set of images of an indoor scene. While classical SfM pipelines rely on a two-step approach where cameras are first estimated using a bundle adjustment in order to ground the ensuing multi-view stereo stage, both our poses and dense reconstructions are a direct output of an altered bundle adjuster. To this end, we parametrize each depth map with a linear combination of a limited number of basis "depth-planes" predicted in a monocular fashion by a deep net. Using a set of high-quality sparse keypoint matches, we optimize over the per-frame linear combinations of depth planes and camera poses to form a geometrically consistent cloud of keypoints. Although our bundle adjustment only considers sparse keypoints, the inferred linear coefficients of the basis planes immediately give us dense depth maps. RidgeSfM is able to collectively align hundreds of frames, which is its main advantage over recent memory-heavy deep alternatives that can align at most 10 frames. Quantitative comparisons reveal performance superior to a state-of-the-art large-scale SfM pipeline.