An Analysis on Intelligent Systems for Remote Sensing Satellite Image Processing and Classification

Abstract

The integration of intelligent systems to the remote sensing satellite image processing and classification has greatly changed. This paper offers a synthesis of the subject, with respect to intelligent systems’ contribution to the improvement of these processes’ accuracy and speed. Accuracy of different methods such as machine learning algorithms, artificial neural networks, and deep learning techniques in the extraction of information from satellite image comprehension is considered a research interest. The presented problems and open issues are data complexity, feature extraction, and classification accuracy over the data, along with new methods in enhancing the intelligent systems to minimize those problems. It expands knowledge of intelligent systems’ contribution to remote sensing applications by outlining how these advancements have influenced the progression of image analysis for the given research goals. This research work gives the summary of our research by outlining the techniques used in the study, the problems solved, and the general outcomes of incorporating intelligent systems in the area of remote sensing and satellite image analysis. Accuracy analysis results for the SVM based methodology with spatial-spectral features include 90% of accuracy, 88% of the precision, and 90% of the F1-score, which in turn makes it easy to make sound decisions when using satellite imagery in different fields like agriculture, urban development, and environment.