Three-dimensional rapid registration and reconstruction of multi-view rigid objects based on end-to-end deep surface model.
Shengzan YanLijun XuShushan WangPublished in: J. Supercomput. (2020)
Keyphrases
- surface model
- multi view
- rigid objects
- three dimensional
- dynamic scenes
- surface reconstruction
- image sequences
- single view
- multiple views
- depth map
- point correspondences
- articulated objects
- motion analysis
- photometric stereo
- camera calibration
- rigid body
- d objects
- surface representation
- object boundaries
- motion parameters
- scene reconstruction
- range images
- computer vision
- high quality
- visual hull
- semi supervised
- structure from motion
- optical flow
- active learning
- least squares
- medical images
- viewpoint
- reconstruction method
- point cloud
- moving objects
- object recognition
- stereo matching
- shape model
- camera motion