A segmentation-based multi-scale framework for the classification of epithelial and stromal tissues in H&E images.
Filiz BunyakAdel HafianeZahraa Al-MilajiIlker ErsoyAnoop HaridasKannappan PalaniappanPublished in: BIBM (2015)
Keyphrases
- multiscale
- image analysis
- edge detection
- microscopic images
- segmentation algorithm
- image classification
- region based segmentation
- accurate segmentation
- pigmented skin lesions
- multiple scales
- image segmentation
- segmentation errors
- segmentation method
- test images
- image regions
- variational framework
- skin lesion
- input image
- fully automatic
- segmentation accuracy
- level set
- adaptive thresholding
- automatic segmentation
- medical images
- cell nuclei
- segmentation scheme
- unsupervised segmentation
- image processing
- microscopy images
- feature extraction
- image registration
- grey level co occurrence matrix
- prior shape knowledge
- brain mr images
- image data
- ground truth
- image features
- shape prior
- image representation
- pixel level
- segmented images
- natural images
- gabor filters
- region growing
- image retrieval
- cancer diagnosis
- texture segmentation
- feature vectors
- medical imaging
- mr images
- object recognition
- multiple objects
- feature selection
- pixel classification
- scale space