Integration of Gibbs Markov Random Field and Hopfield-Type Neural Networks for Unsupervised Change Detection in Remotely Sensed Multitemporal Images.
Ashish GhoshBadri Narayan SubudhiLorenzo BruzzonePublished in: IEEE Trans. Image Process. (2013)
Keyphrases
- markov random field
- remotely sensed
- change detection
- remotely sensed images
- mrf model
- satellite images
- random fields
- remote sensing images
- remote sensing
- neural network
- multispectral
- higher order
- map estimation
- graph cuts
- unsupervised change detection
- maximum a posteriori
- mrf models
- pairwise
- energy function
- image segmentation
- image registration
- belief propagation
- parameter estimation
- neighboring pixels
- image data
- satellite imagery
- input image
- maximum a posteriori probability
- land cover
- gibbs distribution
- image features
- image analysis
- prior model
- smoothness prior
- data streams
- hyperspectral images
- multispectral images
- high resolution
- infrared
- spatial information
- urban areas
- motion estimation
- natural images