Unsupervised variational classification through image multi-thresholding.
Luc KlaineBenoît VozelKacem ChehdiPublished in: EUSIPCO (2005)
Keyphrases
- image segmentation
- image classification
- edge detection
- image content
- input image
- image data
- image features
- supervised learning
- post processing
- image analysis
- pattern recognition
- support vector
- multiscale
- segmentation method
- image feature extraction
- classification accuracy
- unsupervised learning
- preprocessing stage
- thresholding method
- semi supervised
- image representation
- single image
- adaptive thresholding
- decision trees
- segmentation algorithm
- test images
- supervised classification
- hyperspectral images
- image processing
- text classification
- denoising
- high resolution
- image retrieval
- feature extraction
- feature selection
- thematic mapper
- enhanced image
- textural features
- extracted features
- energy function
- markov random field
- optical flow
- feature space