Hybrid patch-based and image-wide classification of confocal laser endomicroscopy images in Barrett's esophagus surveillance.
Elisa VeroneseEnrico GrisanGiorgio DiamantisGiorgio BattagliaCristiano CrostaCristina TrovatoPublished in: ISBI (2013)
Keyphrases
- image classification
- input image
- image data
- image collections
- image features
- test images
- feature based registration
- image analysis
- image regions
- image retrieval
- thematic mapper
- preprocessing stage
- region of interest
- image pixels
- segmentation algorithm
- segmentation method
- image content
- laser scanning
- image set
- image matching
- pixel values
- edge detection
- biomedical images
- low resolution images
- microscope images
- automatic segmentation
- human retina
- skin lesion
- image representation
- low level visual features
- lighting conditions
- feature points
- image structure
- natural images
- visual features
- feature vectors
- three dimensional
- single image
- visual inspection
- contrast enhancement
- image super resolution
- high resolution
- image database
- low resolution
- feature extraction
- image patches
- object recognition
- low contrast
- ct scans
- image registration
- computer vision
- contour extraction
- prior shape knowledge
- feature space
- image quality
- image segmentation
- active contours
- texture synthesis