Geometric Camera Pose Refinement with Learned Depth Maps.
Nathan PiascoDésiré SidibéCédric DemonceauxValérie Gouet-BrunetPublished in: ICIP (2019)
Keyphrases
- depth map
- camera pose
- multi view
- single view
- depth information
- bundle adjustment
- d scene
- reference images
- geometric constraints
- feature points
- camera parameters
- view synthesis
- image sequences
- position and orientation
- structure from motion
- high quality
- stereo matching
- camera calibration
- three dimensional
- camera motion
- pose estimation
- real scenes
- geometric information
- focal length
- point correspondences
- high resolution
- scene reconstruction
- urban environments
- input image
- multiple cameras
- low resolution
- depth estimation
- super resolution
- multiple views
- depth images
- image processing
- epipolar geometry
- stereo vision
- line segments
- input data
- computer vision