Comparison of sampling strategies for object-based classification of urban vegetation from Very High Resolution satellite images.
Simon RougierAnne PuissantAndré StumpfNicolas LachichePublished in: Int. J. Appl. Earth Obs. Geoinformation (2016)
Keyphrases
- satellite images
- landsat tm
- urban areas
- remotely sensed
- sampling strategies
- remote sensing
- change detection
- land cover
- multispectral
- satellite imagery
- remote sensing images
- high resolution satellite images
- pattern recognition
- high resolution
- remotely sensed images
- sampling strategy
- land use change
- hyperspectral
- feature extraction
- decision trees
- multi class
- satellite data
- machine learning
- supervised classification
- class imbalance
- classification algorithm
- training set
- feature space