Scene flow by tracking in intensity and depth data

Abstract

Abstract

Dense Semi-rigid Scene Flow Estimation from RGBD Images

“…Similar to [16], this method suffers from the early linearization of the constancy constraints and from over-smoothing along motion boundaries because of the L 2 -regularization. Quiroga et al [12] define a 2D warping function to couple image motion and 3D motion, allowing for a joint local constraint of the scene flow on intensity and depth data. Although the method is able to deal with large displacements, it fails on untextured regions and more complex motions, such as rotations.…”

“…The work by Herbst [6] follows this idea, but as [16], it lacks a coupling between optical and range flows, and the regularization is done on the optical flow rather than on the scene flow. In [13], a variational extension of [12] is presented. A weighted TV is applied on each component of the 3D motion field, aiming to preserve motion discontinuities along depth edges.…”

Accurate and Computationally-inexpensive Recovery of Ego-Motion using Optical Flow and Range Flow with Extended Temporal Support

“…The benefits of combining the optical flow constraint with range flow have already been recognised [3,10,13]. Barron and Spies embed these as error terms in a global energy functional with additional local first-order smoothness constraints to extract the 3D displacement field [3].…”

“…Barron and Spies embed these as error terms in a global energy functional with additional local first-order smoothness constraints to extract the 3D displacement field [3]. Quiroga et al generate 3D translation fields for image patches using a template matching approach [13], while Haville et al recover the 3D pose changes of pre-segmented image regions using an affine projection assumption [10].…”

Local/global scene flow estimation

“…We combine local and global constraints to solve for the scene flow in a variational framework. Inspired by [4] we locally constrain the scene flow in the image domain but instead of solving for a local motion we get a dense scene flow by performing an adaptive TV regularization. This way, we are able to estimate an accurate dense scene flow while preserving motion discontinuities.…”