Automation of "ground truth" annotation for multi-view RGB-D object instance recognition datasets.
Aitor AldomaThomas FaulhammerMarkus VinczePublished in: IROS (2014)
Keyphrases
- multi view
- d objects
- object recognition
- ground truth
- multiple views
- single view
- viewpoint
- three dimensional
- camera calibration
- depth map
- pose estimation
- shape descriptors
- surface reconstruction
- feature extraction
- range images
- photometric stereo
- color images
- camera parameters
- range data
- object classes
- multi views
- computer vision
- d mesh
- high quality
- view synthesis
- visual hull
- image synthesis