Photographs taken by tyrants like Stalin, Mao, Hitler, Mussolini, and others were frequently altered to reflect the messages of the dictators. They changed backgrounds, inserted people who weren’t there, and removed people who were. They understood that they might alter history if they altered the visual record. Image editing used to take hours in the darkroom, but nowadays, you can do it with a computer and mouse. Fake images may be found everywhere because they are so simple to produce—in supermarket tabloids, fashion magazines, political advertisements, and on social media. How can we determine whether an image is real or not? Hany Farid provides a clear and understandable overview of methods for spotting altered and fraudulent pictures in photos and other digital assets in the MIT Press Essential Knowledge series volume.
Farid has spent two decades creating methods for certifying digital photos. He is a photo forensics expert. These methods simulate the complete image-creation process to identify the digital disturbance caused by image alteration. Each section of the book covers the techniques for evaluating an image, starting with those requiring the least amount of technical knowledge and progressing to those demanding intermediate and advanced levels. Among other things, there are methods for reverse image searches, metadata analysis, spotting image flaws brought on by JPEG compression, image cloning, tracing pixel patterns, and spotting computer-generated images. Farid outlines the methods, clarifies when they should be used, and provides image analysis examples in each section.