This study proposes the design and evaluation of a deep learning model using YOLOv8, an advanced object detection algorithm, for object detection and counting in satellite images. The model will leverage techniques like Laplacian of Gaussian (LoG), Difference of Gaussians (DoG), and Determinant of Hessian (DoH) for pre-processing and feature extraction. It will be trained on a curated dataset of satellite images with ground truth annotations for object locations and counts. The evaluation phase will assess the model's accuracy in object detection and counting using metrics like precision, recall, and mean average precision (mAP), as well as its computational efficiency and scalability. The successful development of this model aims to automate object detection and counting tasks in satellite imagery for applications in urban planning, environmental monitoring, and disaster management.
https://doi.org/10.1109/ICDECS59733.2023.10503432
For support, email [email protected]