Picking Out Digital Image Forgeries
Micah Kimo Johnson has developed tools that can help forensic analysts detect digital image forgeries. On October 6, Johnson gave a presentation entitled “Lighting and Optical Tools for Digital Image Forensics.”
The three techniques he described were illumination direction, specularity, and chromatic aberration. Illumination direction analyzes light sources in a photograph, using a mathematical approach devised by Johnson. The system can calculate the angle of incident light based on the shadows in a picture and recognize any inconsistencies. This software has been successfully built and tested.
The specularity tool he is working on looks at reflective highlights in images. The example used to display this system was a picture from “American Idol,” in which two contestants had been digitally imposed. He showed that the reflective parts of the photo, such as the eyes, revealed a single light source in the eyes of some people pictured and two light sources in others. The algorithm and program are still in the works for this technology.
Finally, chromatic aberration uses the principles of a camera lens and Snell’s law. The tool examines the natural distortion of a picture caused by a camera lens. If this distortion is not consistent throughout, then the image is most likely forged. Johnson is still perfecting this technology.
While none of these tools is 100 percent effective on its own, when the three are used in concert with forensic analysis, they contribute a great deal to investigations and verifications of forged images.
Abstracted by the National Law Enforcement and Corrections Technology Center (NLECTC) from Network World (10/17/06); Kabay, M.E.
