Noname manuscript No. (will be inserted by the editor) Passive Detection of Image Forgery using DCT and Local Binary Pattern Amani Alahmadi 1 , Muhammad Hussain 1,a , Hatim Aboalsamh 1 , Ghulam Muhammad 1 , George Bebis 2 , Hassan Mathkour 1 Received: date / Accepted: date Abstract With the development of easy-to-use and sophisticated image editing software, the alteration of the contents of digital images has become very easy to do and hard to detect. A digital image is a very rich source of information and can capture any event perfectly, but because of this reason, its authenticity is questionable. In this paper, a novel passive image forgery detection method is proposed based on Local Binary Pattern (LBP) and Discrete Cosine Transform (DCT) to detect copy-move and splicing forgeries. First, from the chrominance component of the input image, discriminative localized features are extracted by ap- plying 2D DCT in LBP space. Then, support vector machine (SVM) is used for detection. Experiments car- ried out on three image forgery benchmark datasets demonstrate the superiority of the method over recent methods in terms of detection accuracy. Keywords Copy-move forgery, Image splicing, Forgery detection, Image forensics, LBP, DCT, SVM 1 Introduction In today’s visual world, digital images have become an integral part of our everyday life due to their ability to convey a wide range of information in a compact way and the availability of digital image acquisition tools. On the other hand, one needs not to be skillful to alter the contents of a digital image without leaving obvi- ous traces of changes because of the development of user-friendly image editing tools. It has become easy 1 College of Computer and Information Sciences, King Saud University, Riyadh, Saudi Arabia 2 Department of Computer Science and Engineering, Univer- sity of Nevada at Reno, USA a Email: mhussain@ksu.edu.sa to use digital images for nefarious designs and nega- tive propaganda on social and electronic media, and hiding the facts, which can be crucial for criminal in- vestigation, medical imaging, scientific discoveries, etc. As such, the authenticity of digital images cannot be taken for granted. Image splicing and copy-move are two very harmful and commonly used types of forgery. Some techniques have already been proposed to detect such forgeries [1]. These techniques are either intrusive (active) or non- intrusive (passive, blind) [2]. An active technique de- tects tampering by verifying the integrity of a signature (embedded by a digital camera) such as watermark; it has a restricted scope due to the limitations of most of the cameras to embed such signatures [1]. On the other hand, a non-intrusive technique has a widespread scope since it depends only on analyzing the characteristics of a digital image[3]. Passive techniques can be broadly classified into learn- ing based [4] and block-matching based methods [5][6]. The latter category of methods detects forgery by local- izing the regions, which have been tampered by copy- paste. It is useful for sensitive applications like evidence in court rooms, insurance claims, etc., but it is time- consuming and unsuitable for applications like social media, where a bulk of images is being shared every day, and it is enough to verify whether an image is forged or not. In this paper, we propose a learning based passive technique that detects copy-move and image splicing forgeries. The challenge in a learning based method is how to model the change incurred by tampering. The key idea of the proposed method was inspired from an- alyzing the tampering procedure. When tampering is done, it disturbs the local distribution of micro-edge patterns by introducing new micro-patterns in the in- terior of the pasted region and sharp edges along its