DENet: Double-Encoder Network With Feature Refinement and Region Adaption for Terrain Segmentation in PolSAR Images.
Xuan ZengZhirui WangXian SunZhonghan ChangXin GaoPublished in: IEEE Trans. Geosci. Remote. Sens. (2022)
Keyphrases
- segmented regions
- image regions
- segmentation method
- segmented images
- grey level
- segmentation algorithm
- input image
- image features
- test images
- three dimensional
- image analysis
- region segmentation
- region of interest
- image segmentation
- image pixels
- homogeneous regions
- level set framework
- segmentation result
- segmentation scheme
- edge detection
- ground truth
- image data
- microscopy images
- pixel level
- pixel wise
- initial segmentation
- image segments
- intensity distribution
- object regions
- accurate segmentation
- foreground and background
- medical images
- gradient information
- intensity variations
- level set
- salient regions
- image retrieval
- optimal segmentation
- image structure
- region growing
- primitive features
- segmentation errors
- gray level co occurrence matrix
- texture and shape features
- watershed segmentation
- segmentation accuracy
- multiple objects
- shape prior
- watershed transform
- textural features
- superpixels
- watershed transformation
- region features
- adjacent regions
- bit rate
- feature vectors