Feature- and depth-supported modified total variation optical flow for 3D motion field estimation in real scenes.
Thomas MüllerJens RannacherClemens RabeUwe FrankePublished in: CVPR (2011)
Keyphrases
- real scenes
- total variation
- motion field estimation
- optical flow
- depth map
- image restoration
- denoising
- image denoising
- image sequences
- motion field
- augmented reality
- motion estimation
- image processing
- scene geometry
- super resolution
- depth information
- high quality
- motion model
- optical flow estimation
- vector field
- stereo matching
- stereo images
- flow field
- stereo pair
- motion segmentation
- feature vectors
- camera motion
- multi view
- image features
- three dimensional
- moving objects
- feature space
- image matching
- machine learning
- image data
- input image