Detection of Seam-Carving Image Based on Benford's Law for Forensic Applications

Detection of Seam-Carving Image Based on Benford's Law for Forensic Applications

Guorui Sheng (School on Information and Electronical Engineering, Ludong University, Yantai, China) and Tiegang Gao (College of Software, Information Security Technology Lab, Nankai University, Tianjin, China)
Copyright: © 2016 |Pages: 11
DOI: 10.4018/IJDCF.2016010104
OnDemand PDF Download:
No Current Special Offers


Seam-Carving is widely used for content-aware image resizing. To cope with the digital image forgery caused by Seam-Carving, a new detecting algorithm based on Benford's law is presented. The algorithm utilize the probabilities of the first digits of quantized DCT coefficients from individual AC modes to detect Seam-Carving images. The experimental result shows that the performance of proposed method is better than that of the method based on traditional Markov features and other existing methods.
Article Preview

1. Introduction

With the fast development of digital image processing techniques, more and more image processing software like GIMP and Adobe Photoshop appeared and provided more and more powerful abilities to process digital images. As such software affording the convenience of processing digital image, they also generate the possibility of image tampering and misusing. If tampered images are used in the news media, scientific discovery, insurance and the court evidence, the social order and living order will be destroyed. False image change the true face of the world, even influence the judicial justice. Therefore, it becomes more and more important to confirm the authenticity of an image. The image forensics technique offers a favorable means for image tampering detection.

The image forensic techniques can be divided into two categories: proactive forensics and passive forensics. For proactive forensics, special information such as watermark is embedded into original image. When checking the authenticity of an image, such information will be extracted and compared with original embedded information to draw the conclusion. The main technique of proactive forensics is watermarking (Reyc, 2003; Potdar, 2005; Weng, 2013; Liu, 2010), which takes advantage of redundant information in digital images. Nowadays, some digital cameras support embedding watermark into digital image automatically when taking pictures. The main limitation of this kind of forensic is that, when facing large amounts of digital images, people usually can’t know which image should be checked in advance. On the contrary, the passive forensics can judge the authenticity of an image without any embedded information in advance. The wide application of this kind of technique motivated many researchers focus on this domain.

The main image tamping methods include copy-move, splicing, resampling and so on. Copy-move means copying one part of the original image and then pastes it to the other part of the same image. This kind of method is usually for the purpose of hiding or duplicating special contents such as a special object or people. One of the main corresponding forensic methods aiming at copy-move tamping is to find if there are homogeneous regions in the image, which can be achieved by using characteristics in frequency domain of the image (Fridrich, 2003; Amerini, 2011; Amerini, 2013).. Image splicing means compounding two or more images into one, the corresponding forensic methods aiming at this respect are mainly achieved with the help of analysis of image features such as high order moment and bicoherence spectrum (Shi, 2007; Fu, 2006). Image resampling includes operations such as resizing, rotating and scaling. This kind of operation will resample the original image and cause periodic correlation which can be detected for forensic purpose.

Complete Article List

Search this Journal:
Volume 14: 1 Issue (2022): Forthcoming, Available for Pre-Order
Volume 13: 6 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing