Title
Hierarchical segmentation of urban satellite imagery.
Abstract
•We extract the features of the urban area satellite images using Gabor wavelet.•Gabor wavelet applied in eight different directions (phases) considering sparsity and no need to add any sparse coding techniques.•Categorization and segmentation of satellite images has done hierarchically using Bayesian classifier.•Relaxation labeling eliminates unclassified and misclassified parts that it improves the accuracy rate.•There is a combination among contextual, structural, spectral, and statistical information toward classification.
Year
DOI
Venue
2014
10.1016/j.jag.2014.01.014
International Journal of Applied Earth Observation and Geoinformation
Keywords
Field
DocType
Very high resolution satellite imagery,Gabor wavelet,Bayesian classifier,Relaxation labeling
Categorization,Relaxation labeling,Satellite imagery,Pattern recognition,Naive Bayes classifier,Gabor wavelet,Segmentation,Artificial intelligence,Novelty,Geography,False positive paradox
Journal
Volume
ISSN
Citations 
30
0303-2434
1
PageRank 
References 
Authors
0.37
21
4