Unsupervised learning segmentation for dynamic speckle activity images.
Lucía Isabel PassoniAna Lucía Dai PraGustavo J. MeschinoMarcelo GuzmánChristian WeberHéctor RabalMarcelo TriviPublished in: CoRR (2014)
Keyphrases
- medical ultrasound
- unsupervised learning
- segmentation algorithm
- image analysis
- edge detection
- segmentation method
- image data
- image regions
- cardiac ultrasound images
- segmentation errors
- image database
- fully automatic
- test images
- microscopic images
- accurate segmentation
- segmented images
- segmentation accuracy
- adaptive thresholding
- speckle noise
- low depth of field
- image segmentation
- ground truth
- ultrasound images
- image slices
- grey level
- image retrieval
- object recognition
- pixel level
- pixel wise
- input image
- microscopy images
- positron emission tomography
- image features
- sar images
- automatically segmented
- image registration
- image segments
- tubular structures
- cell nuclei
- supervised learning
- multiple objects
- machine learning
- multiplicative noise
- level set
- noise reduction
- pigmented skin lesions