Assessing land cover for Bahar Al-Najaf using maximum likelihood (ML) and artificial neural network (ANN) algorithms

Change information of the earth’s surface is becoming more and more important in monitoring the local, regional and global resources and environment. In this paper, two algorithms maximum likelihood (ML) and artificial neural network (ANN) have been applied for a change detection at the Baher Al-Najaf from 2016 to 2020 and using two satellite imagery Landsat and sentinel. The accuracy assessment of these algorithms was based on the confusion matrix such as overall accuracy, and the kappa index were calculated for each created map. It was found (ANN) classifier is the better than (ML) classifier. As well as two different experiments were conducted to analyze the network depth and optimization of ANN classifier, and the results showed that the proposed classifier achieves higher overall accuracy and kappa index with pansharpening image. Finally, this study proved (ANN) classifier ability to extract useful high-level features in the classification process.