Semantic Segmentation Based Copy Paste Forgery Detection

Abstract

Semantic segmentation refers to the process of finding meaningful sections in a picture. Semantic segmentation is employed in a variety of domains, including the military, robotics, and biomedicine. A new copy-paste forgery approach based on semantic segmentation is provided in this research. Forgery zones were found with improved accuracy and in less time using discrete cosine transforms and semantic segmentation. Because the suggested technique uses semantic segmentation, each region is compared inside its own class, resulting in both time savings and increased performance.