A probabilistic approach to segmentation and classification of neoplasia in uterine cervix images using color and geometric features.
Yeshwanth SrinivasanDana HernesBhakti TulpuleShuyu YangJiangling GuoSunanda MitraSriraja YagneswaranBrian NutterJose JeronimoBenny PhillipsL. Rodney LongDaron FerrisPublished in: Medical Imaging: Image Processing (2005)
Keyphrases
- geometric features
- skin lesion
- image regions
- image classification
- segmentation algorithm
- segmentation method
- image analysis
- salient regions
- illumination correction
- geometric structure
- image features
- classification accuracy
- shape features
- image data
- input image
- feature vectors
- level set
- cervical cancer
- shape descriptors
- complex background
- image registration
- image segmentation
- feature extraction
- computer vision
- multiscale
- feature space
- salient features
- shape prior
- image matching
- matching algorithm
- spatial information
- low level
- active contours