Scanline Resolution-Invariant Depth Completion Using a Single Image and Sparse LiDAR Point Cloud.
Kwonyoung RyuKang-il LeeJegyeong ChoKuk-Jin YoonPublished in: IEEE Robotics Autom. Lett. (2021)
Keyphrases
- single image
- point cloud
- surface reconstruction
- structure from motion
- stereo matching
- multiple images
- shape recovery
- depth map
- d scene
- stereo pair
- point sets
- high resolution
- lidar data
- super resolution
- shape from shading
- depth information
- point cloud data
- image pairs
- stereo camera
- dynamic programming
- stereo correspondence
- laser scanner
- sparse representation
- object surface
- high dimensional
- low resolution
- intrinsic images
- light source
- hough transform
- vanishing points
- object recognition
- video sequences