MSc Thesis in Computer Science discusses Classification of Composite Gap Radar Images
MSc. thesis in the Dept. of Computer Science at the UOT discussed the classification of composite gap radar using multiple segmentation and a hierarchical classification structure.
The study presented by student Aseel Sami Ali/ information systems, proposed a classification system based on the transmission of education with bypass neural networks to classify the SAR image, and it consists of 8 stages: the stage of importing libraries, and preparing data that transforms the input images into a format that meets the requirements of the previously trained user form. Feature extraction that uses a pre-trained model that acts as a feature extractor on the training images, the image classification stage that feeds the extracted features, the design of a convolutional neural network for training and the production of a coached model, a diagram of the model phase, the evaluation of the model phase, generation of the confusion matrix phase and the test phase on the model trained to predict Category each image tested.
Five proposed experiments were conducted on this system to compare results, and they gave correct predictions for all groups. The discussion was held in Computer Science.
Discussion Committee Consists of :-
- Dr. Abdulameer Abdllah / Head
- Prof. Dr. Ghada Kadhum Nimaa/ Member.
- Prof. Dr. Rahim A. Oglaa/ Member.
- Prof. Dr. Ahmed Alaa Oglaa/ Supervisor.
