Society as a whole has developed a dependence upon digital media, however there is often more to it than meets the eye. Marcel Proust once wrote: “The real voyage of discovery consists not in seeking new landscapes but in having new eyes.” The ubiquity of digital cameras in modern-day life has provided us with new eyes with which we can view the world. However, parallel to the rise in use of digital images is the increase in the sophistication of methods to doctor such images, so can these new eyes be trusted? It is now common to assume that photographs of celebrities or landscapes have been touched up, or “Photoshopped,” in some manner. In many cases, this digital manipulation is innocuous, yet there are scenarios where it is important, or even essential, to verify the origin and accuracy of digital images. As such, digital image forensics is a growing field of research concerned with obtaining quantitative evidence to analyze and authenticate digital images. This book, edited by Sencar and Memon, seeks to provide an introduction and overview of this burgeoning area of research.
The book is split into four parts: the first part deals with the foundations of digital imaging, the second and third parts review state-of-the-art techniques for image source attribution and authenticity verification, and the final part looks at practical image forensics with a particular focus on the courtroom perspective. A cursory glance at the table of contents would imply that a seemingly inordinate amount of space has been given to the introductory chapters. A naive reader may assume that the background can be skipped, so as to quickly move on to the actual heart of the book. However, in the context of digital imaging, and digital image forensics in particular, this would be a grave mistake. Many of the techniques described in Parts 2 and 3 are heavily reliant on a solid understanding of not only the digital imaging pipeline, but also the optics involved in capturing the image and the methods used to digitally store such images. Therefore, Part 1 is essential in laying the foundation for the main part of the book, and it is thankfully done in an easy-to-read and accessible way.
As mentioned above, the so-called heart of this book is Parts 2 and 3, where methods for the forensic analysis of digital images are outlined. A digital image contains not only visual content (that is, the actual image itself), but also metadata, often referred to as multimedia source. The chapters in Parts 2 and 3 deal with how both of these strands can be used in various digital image forensic tasks. These tasks include identifying whether the image has been tampered with or the metadata changed; finding the camera that was used to take a specific picture; and examining the integrity and authenticity of an image, for example, identifying whether the image is real or computer generated. All of these areas are neatly addressed in separate chapters, with very little overlap between them. Often in edited books such as this, where all chapters deal with a specific problem domain, there is a large amount of redundancy, as time is spent explaining and reexplaining similar problem definitions. However, this is not the case here; each chapter does not have a needlessly long and repetitive introduction, but rather focuses on the particular problem at hand. This improves the readability of the book without impeding its use as a reference text.
One drawback of the book is that, since each chapter is a contribution by a different author, there is a lack of critical analysis and common conclusions drawn from all the chapters. Perhaps it is unfair to expect such an analysis from an edited book of separate contributions; however, when multiple chapters deal with similar problems, it can be difficult to ascertain the relative influences, strengths, and weaknesses of certain methods. It is also difficult to reconcile the final part of the book to its predecessors. Part 4 consists of only two chapters and aims to deal with digital image forensics in practice. The chapter by Rebecca Mercuri on digital image forensics in a courtroom setting, while interesting, is perhaps inaccessible to a technical audience and, moreover, may not be overly useful for the average practitioner of digital image forensics. Whilst it is important to see how the techniques described in previous chapters could be used in a courtroom, the chapter does feel substantially different than the others. This is amplified by the fact that the second chapter in Part 4 deals with counter forensics, and it is difficult to see how the two fit together.
These problems aside, this book is an excellent introduction to the field. As a whole, it is well organized and provides sufficient detail to understand the basic techniques and general trends within the area. It also shows that, thanks to digital image forensic techniques, we can begin to trust the “new eyes” that we have been given, albeit with a healthy dose of caution.