Advancements in satellite image classification : methodologies, techniques, approaches and applications

• DOI: 10.1080/01431161.2021.1954261
• Publication Date: 2021-09-21T05:45:45Z
• Authors: G. A. Fotso Kamga; L Bitjoka; T. Akram; A. Mengue Mbom; S. Rameez Naqvi; Y. Bouroubi
• Keywords: Contextual image classification; Disjoint sets; Remote sensing application; Feature (linguistics)

Segmentation and classification are two imperative, yet challenging tasks in image analysis for remote-sensing applications. In the former, an is divided into spatially continuous, disjoint, homogeneous regions, called clusters, terms of their various properties: shape, intensity, texture, colour, contrast, etc. Classification, on other hand, applied later process, to recognize or categorize individual objects targets. Each task plays important role refinement enhancement utilizations remote sensing images. Driven by recent progress earth observation sensor technology, satellite systems earth-observation applications have seen significant growth progress. This has led a notable increase number published materials these areas. We present overview horizons that modern domain promises efficient processing imagery. begin defining sensing, specifically context its potential application areas, highlight importance pre-processing feature extraction steps' accurate classification. Various works proposed novel segmentation, extraction/selection, methods; been collected duly reported this work. The deep learning method given special attention due relatively limited dependence training data, wide spectrum applications, ability autonomously classify images with higher accuracy. conclude presenting critical evaluation contributions domain.