Direkt zum Inhalt springen
Computer Vision Group
TUM Department of Informatics
Technical University of Munich

Technical University of Munich

Menu
Home Research Areas Optical Flow Estimation

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Last revision Both sides next revision
research:optical_flow_estimation [2012/01/20 11:58]
127.0.0.1 external edit
research:optical_flow_estimation [2012/06/08 10:24]
steinbrf
Line 1: Line 1:
 ====== Optical Flow Estimation ====== ====== Optical Flow Estimation ======
 +
 +
 +Estimating the motion of every pixel in a sequence of images is a problem with many applications in computer vision, such as image segmentation,​ object classification,​visual odometry, and driver assistance.
 +
 +In general, optical flow describes a sparse or dense vector field, where a displacement vector is assigned to certain pixel position, that points to where that pixel can be found in another image.
 +In the context of scene flow estimation, which is performed on images with additional depth values, every pixel is assigned a depth displacement as well.
 +
 +Since much of the structural information of a 3D scene gets lost in the imaging process, so does the motion information. The estimation of the "​correct"​ projected motion in an image sequence is therefore highly ill-posed and has to be aided by additional priors such as the regularity of the motion.
 +Our group mainly focuses on optical flow estimation by means of variational methods, that allow a clear formulation of the assumptions incorporated into the estimation process and generally produce dense vector fields.
  
 Contact: [[members:​steinbrf|Frank Steinbrücker]] Contact: [[members:​steinbrf|Frank Steinbrücker]]

Rechte Seite

Informatik IX
Chair of Computer Vision & Artificial Intelligence

Boltzmannstrasse 3
85748 Garching

info@vision.in.tum.de