Scene Flow from Depth and Color Images
By Antoine Letouzey, Benjamin Petit, Edmond Boyer. 2011
46 citations recorded by the source corpus.
In this paper we consider the problem of estimating a 3D motion field using multiple cameras. In particular, we focus on the situation where a depth camera and one or more color cameras are available, a common situation with recent composite sensors such as the Kinect. In this case, geometric information from depth maps can be combined with intensity variations in color images in order to estimate smooth and dense 3D motion fields. We propose a unified framework for this purpose, that can handle both arbitrary large motions and sub-pixel displacements. The estimation is cast as a linear optimization problem that can be solved very efficiently. The novelty with respect to existing scene flow approaches is that it takes advantage of the geometric information provided by the depth camera to define a surface domain over which photometric constraints can be consistently integrated in 3D. Experiments on real and synthetic data provide both qualitative and quantitative results that demonstrate the interest of the approach.
Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Aerospace Engineering
- Carsten Rother
- Daniel Cremers
- Thomas Brox
- Andrew William Fitzgibbon
- Michael Hornacek
- Julian Quiroga
- James L. Crowley
- Frederic Devernay
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Jitendra Malik
- Cordelia Schmid
- Andrew Zisserman
- Ming-Hsuan Yang
- Dacheng Tao
- Andrea Vedaldi
- Trevor Darrell
Citations in the corpus
- A primal-dual framework for real-time dense RGB-D scene flow (2015, methodology): “[12] makes use of a weighted quadratic regularizer and considers a set of sparse feature correspondences to handle large displacements.”
- Scene flow by tracking in intensity and depth data (2012, methodology): “[18] directly estimate the 3D motion field using photometric constraints and a global regularization term.”
- Dense Semi-rigid Scene Flow Estimation from RGBD Images (2014, background): “The coupling issue can be solved as in [10], where the scene flow is directly solved using the depth data to constrain the 3D motion in the image domain.”
- SphereFlow: 6 DoF Scene Flow from RGB-D Pairs (2014, methodology): “[17] claim novelty for using RGB-D data obtained from a Kinect.”