Land Use Classification of Satellite Images With Convolutional Neural Networks (cnns)

Deep-learning neural networks continue to increase in effectiveness and capability. Specifically, Convolutional Neural Networks (CNNs) have proven to offer highly accurate results in differentiating graphical images. Using satellite imagery, CNNs have been highly successful in the classification of land use. This successful classification, however, depends on the model structure, data components, and classification methods utilized. While alternative approaches exist, the complexity and sophistication of such classification methods impose increased computational costs. Effective manipulation of CNN components can offer computational savings and accurate classification of land use in satellite images.