In this paper, we present a method for performing secure image de-duplication at the image level based on the MD5 protocol. We have discovered that the degree of similarity between images directly influences the number of images stored in the cloud. However, it is important to note that our current approach is limited to handling images that are nearly identical, with only slight variations. In future research, we aim to tackle this issue on a broader scale by incorporating additional image operations such as scaling, rotation, cropping, multiple viewpoints, varying lighting conditions, and...