Classification and segmentation of orbital space based objects against terrestrial distractors for the purpose of finding holes in shape from motion 3D reconstruction.
T. Nathan MundhenkArturo FloresHeiko HoffmanPublished in: Intelligent Robots and Computer Vision: Algorithms and Techniques (2014)
Keyphrases
- object shape
- multibody
- object segmentation
- space time
- video objects
- shape prior
- appearance features
- motion segmentation
- moving objects
- geometrical constraints
- maximum likelihood criterion
- relative position
- object contours
- segmentation algorithm
- remote sensed images
- rigid objects
- multiple objects
- figure ground segmentation
- articulated objects
- level set
- image segmentation
- image sequences
- image segments
- medical images
- motion tracking
- target object
- outdoor scenes
- deformable models
- d scene
- computer vision
- structure from motion
- perspective images
- three dimensional
- image measurements
- feature vectors
- optical flow
- shape model
- energy function
- shape representation
- motion model
- object shapes
- rigidly moving
- appearance model