A survey on deep learning-based precise boundary recovery of semantic segmentation for images and point clouds.
Rui ZhangGuangyun LiThomas A. WunderlichLi WangPublished in: Int. J. Appl. Earth Obs. Geoinformation (2021)
Keyphrases
- point cloud
- three dimensional
- image classification
- image features
- weakly supervised
- test images
- structure from motion
- input image
- machine learning
- superpixels
- bounding box
- image annotation
- lighting conditions
- feature points
- segmentation algorithm
- image retrieval
- multiscale
- feature extraction
- region of interest
- similarity measure
- conditional random fields
- spatial information
- image regions
- segmentation method
- object recognition
- high dimensional