SegSALSA-STR: A convex formulation to supervised hyperspectral image segmentation using hidden fields and structure tensor regularization.
Filipe CondessaJosé M. Bioucas-DiasJelena KovacevicPublished in: WHISPERS (2015)
Keyphrases
- structure tensor
- image segmentation
- remote sensing
- optical flow
- computer vision
- anisotropic diffusion
- image processing
- edge information
- image analysis
- learning algorithm
- image data
- markov random field
- segmentation method
- multiscale
- region growing
- segmentation algorithm
- semi supervised
- graph cuts
- supervised learning
- machine learning
- diffusion process
- feature selection
- image sequences
- spectral clustering
- active contours
- similarity measure
- level set
- edge detection
- high quality