Detection of ships in sea regions using remote sensing images (satellite images) is an important task for ensuring maritime surveillance and security and addressed in this work. However, satellite images often have low spatial resolution, which makes it difficult to accurately detect and classify ships. For enhancing low spatial resolution images into high spatial resolution images, super-resolution techniques have been implemented so that the detection and classification of ships becomes an easy task. In this work, a ship detection method is proposed that uses SwinIR and ESRGAN to increase the resolution of the image, and ships are detected from this high-resolution image using CNN and U-Net algorithms. Three performance metrics are used to evaluate the outcome of the work, namely structural similarity index (SSIM), normalized mutual information (NMI), and mean square error (MSE). The proposed model is very useful for ship search and rescue operations and can also be used by the military to watch over vast sea areas.