SA-MVSNet: Self-attention-based multi-view stereo network for 3D reconstruction of images with weak texture.
Ronghao YangWang MiaoZhenxin ZhangZhenlong LiuMubai LiBin LinPublished in: Eng. Appl. Artif. Intell. (2024)
Keyphrases
- multi view stereo
- multi view
- structure from motion
- three dimensional
- urban environments
- stereo reconstruction
- depth map
- point cloud
- bundle adjustment
- ray tracing
- image retrieval
- texture mapping
- computer vision
- image features
- depth estimation
- high fidelity
- input image
- camera parameters
- surface reconstruction
- stereo vision
- stereo algorithms
- computer graphics