A master's thesis at the College of Computer Science and Information Technology at University of Basrah explored the development of a computer vision model for the accurate detection of objects in underwater environments.
The thesis, submitted by student Shaimaa Khader Hussein, aimed to improve object detection using modern computer vision techniques by developing the YOLOv8 model through the integration of the ASPP (Atrous Spatial Pyramid Pooling) module to enhance multi-scale feature representation.
The study concluded that the proposed model achieves a significant improvement in detection accuracy while maintaining time efficiency, confirming its potential for effective application in underwater imaging and field research.
