The Department of Computer Science held a PhD dissertation defense

Ph.D student “Ihsan Ali Karim defended his dissertation entitled “Detection of Sand Dunes using Multispectral Satellite Images based on Machine Learning Techniques”. The study dealt with machine learning techniques in analyzing satellite images of multispectral satellites in discovering and distinguishing terrestrial phenomena. The study focused on the phenomenon of sand dunes because of their importance and environmental impact in Iraq based on Landsat-9 satellite (Baiji and Najaf-Samawah fields a case study). Nine methods were used in this field. The study concluded that the common LibSVM method was one of the least efficient and accurate methods and required more time in the analysis. Also, linear equations were found capable of distinguishing and sorting the sand dune phenomenon with efficiency and accuracy exceeding 80% compared to the previously studied methods. The linear equations also proved efficient on other multispectral satellite images (Sentinel-2 and Lansat-5). The defense was accepted with
success.