Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Photo forensics
Farid H., The MIT Press, Cambridge, MA, 2016. 336 pp. Type: Book (978-0-262035-34-7)
Date Reviewed: Aug 31 2017

From a number of perspectives, Photo forensics is a wonderful book. From the most material point of view, it is simply beautiful. You may even see it on coffee tables. The cover is smartly designed, the paper is of high quality, and the images included are enriched color plates, which incorporate shades and effects. Second, the book is intelligently divided into self-contained chapters, starting with an introduction, fundamental concepts (physics and geometric constraints in the next two chapters), devices (optics and sensors, the next two chapters), storage (bits and pixels, chapters 6 and 7), and applications (chapter 8). Chapter 9 covers higher mathematical concepts that are used throughout the book.

The author introduces the field with an interesting anecdote about why he became interested in it. Fortuitously, he saw a legal book describing the rules of evidence, which discussed the definition of an original recording or photograph. The author was puzzled by the vague criterion of being “readable by sight” [1]. This led to a multi-year search of how one could possibly authenticate a photograph, obtained conventionally or digitally, culminating in the concept of correlating neighboring pixels. Regarding this, see section 7.3 on resampling methods, where the author augmented the size of a fish caught at sea by many-fold and the picture was perceptually accurate but left interpolation artifacts in neighboring pixels, which would not arise by chance. There, a computational algorithm is provided to detect such anomalies.

Physics-based constraints are due to shadows, lighting, and reflection (chapter 2). Geometric constraints come about because of vanishing points, measuring distances in the image plane, and overall 3D modeling (chapter 3). Optical distortions (chapter 4) can occur when the lens bends the light and curves lines where they should be straight; lens flare can obscure parts of the image data. Flare is caused when light enters into the image that is not necessary for the formation of the image.

Chapter 5 handles sensor and fixed-pattern noise where certain pixels tend to have brighter intensity than provided by the background. The latter is measured by a photo response nonuniformity (PRNU) measure. Because of the ever-increasing number of constraints that technology introduces to image formation, accurate photo forensic techniques are all the more important. Indeed, the author forewarns law enforcement professionals to work diligently when applying forensic techniques to analyze image data. They need to check over their calculations because other people’s lives are affected by any mistakes.

Chapter 8 is a very interesting chapter. It provides case studies that the author worked on. The first is a famous hoax perpetrated by Montreal students of an eagle snatching a child [2]. It has garnered more than 45 million views on YouTube. Finding a frame that has clear cast shadows (section 2.1) of the child, eagle, and background elements enabled the author to uncover the inconsistent assumptions of where the source of light came from. The students later published a notice explaining that they had only produced it for a digital photography class.

The second test case was an image of a dwarfed soccer player, which Reuters [3] had been convinced was real. The author determines that in fact, this optical illusion is entirely plausible due to the degenerate point of view where the player is leaning back in the direction of camera alignment. The last case study is that of a picture of Lee Harvey Oswald holding a shotgun in the backyard. Many considered this image fallacious due to implausible lighting. The author’s in-depth analysis showed that in fact it is a consistent image and can be considered authentic.

The author does a great job in presenting the field of photo forensics in a manner that is accessible to undergraduate, graduate, and professional students alike. This book is a must-read for anyone interested in the field.

Reviewer:  Minette Carl Review #: CR145520 (1711-0718)
1) Rule 1001. Definitions that apply to this article. Legal Information Institute. https://www.law.cornell.edu/rules/fre/rule_1001 (06/13/2017).
2) Golden eagle snatches kid. YouTube, Dec. 18, 2012, posted by MrNuclearCat, https://www.youtube.com/watch?v=CE0Q904gtMI (06/13/2017).
3) Footballer gets dwarfed in optical illusion photo. Reuters. http://pictures.reuters.com/Doc/RTR/Media/TR3/8/a/6/2/RTX12LGS.jpg (04/26/2017).
Bookmark and Share
  Reviewer Selected
Editor Recommended
 
 
Applications (I.4.9 )
 
 
General (I.4.0 )
 
 
Security and Protection (K.6.5 )
 
Would you recommend this review?
yes
no
Other reviews under "Applications": Date
The Abingdon Cross Benchmark Survey
Kendall J. Computer 22(7): 9-18, 1989. Type: Article
Jul 1 1990
The Cosmic Worm
Goldman J., Roy T. IEEE Computer Graphics and Applications 14(4): 12-14, 1994. Type: Article
Jul 1 1995
A multiscale elastic registration scheme for retinal angiograms
Nunes J., Bouaoune Y., Delechelle E., Bunel P. Computer Vision and Image Understanding 95(2): 129-149, 2004. Type: Article
Jan 11 2005
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy