An Efficient Method for Image Forgery Detection Based on MEHO-SAE Model

Nowadays, more and more people can afford to take images, making photography a popular pastime. The wealth of information stored in images makes them indispensable in our daily lives. In order to get more information out of images, we usually have to edit them. Many methods exist to enhance photographs; however, these same techniques could also be utilized to create misleading images, resulting in the dissemination of inaccurate information. Already a major worry, image forgeries are getting even worse as a consequence of this. It is critical to follow the precise order of operations when preprocessing, extracting features, and training the model. The preprocessing includes tasks such as key point extraction and HSV conversion. Feature extraction, the process of obtaining pertinent information from an input picture, is the initial step in developing an object identification system. Using a MEHO-SAE model, the model was trained. This approach appears to be fairly novel when compared to the state-of-the-art MEHO and SAE. The results improved significantly and achieved a level of 97.23 percent accuracy.