Measurement of Copy-Move Forensic Image Similarity Using Distance Function

In digital imagery era, good-editing software allows users to process digital images in an easy way. As a result, image fraud practice is widespread inevitably. Therefore, an image-fraud-detection tool is required to prove the authenticity of a digital image. The growth of digital image manipulation has prompted writers in forensic imagery field to reveal the authenticity of images. In this research the original data images include two images object: beach.jpg and sailing.jpg. Then, manipulated images created manually: beach1.jpg and sailing1.jpg. The experiment is done through several stages of the conversion of color images into the form of grayscale degree followed by determining pattern through histogram of pixel image values. The image similarity is based on calculating distance value of two altered image patterns. Some of the distance functions performed in this research namely Euclidean distance, Manhattan distance and Canberra distance. While image manipulation used are limited into copy-move image manipulation. The Research results that distance function from two different images namely original image and manipulate image compare with original image and manipulated image’s difference object. The experimental result of three distance functions show same result that smallest calculation value is distance calculation of original image and manipulated image from the same object.