The activity of detecting and classifying ships is essential for both the national defense and the nation’s marine security. Many different image processing algorithms have been developed to detect a ship in remote sensing photo as a result of the availability of enormous amounts of high-quality remote sensing images. However, thus the majority of these systems have challenges in phrase of their accuracy, performance as well as level of complexity. In this study, we propose making use of YOLOv5 with the aim to detect ships in the same time using optical remote sensing photos with improving the accuracy of ship detection, performance, and level of complexity by 5% of existing system which means that our detection accuracy is 89.4% for YOLOv5 Model.