Image Spectral Data Classification Using Pixel-Purity Kernel Graph Cuts and Support Vector Machines: A Case Study of Vegetation Identification in Indian Pine Experimental Area.
Mengfei WangWeijie JiaQingjie LiuFengxian MiaoPublished in: IGARSS (2019)
Keyphrases
- support vector
- graph cuts
- energy function
- spatial coherence
- input image
- image segmentation
- labeling problems
- segmentation algorithm
- image classification
- spectral data
- graph cut segmentation
- support vector machine
- kernel function
- hyperspectral images
- classification accuracy
- markov random field
- energy minimization
- cross validation
- image data
- image analysis
- hyperspectral data
- principal components
- satellite images
- svm classifier
- remote sensing
- object segmentation
- feature space
- feature extraction
- feature selection
- kernel methods
- high resolution
- hyperspectral
- generalization ability
- active contours
- machine learning
- multispectral images
- mean shift
- similarity measure
- hyperspectral imagery
- feature vectors