DeGrade It

As submitted by the proposer: Extremely low-quality images – on the order of 10 to 20 pixels in size – appear with frustrating frequency in many investigations. Even the most advanced de-noising and super-resolution technologies are unable to extract useful information from such low-quality images.

However, useful information is present in such highly degraded images. We propose that the way to extract this information is not to enhance an image, but to degrade an image. In particular the researcher proposes a new forensic approach that searches the space of possible content that when degraded matches the content of the low-quality image in question. Such an approach is particularly well suited to deciphering the identity of a license plate or the identity of individual in an extremely low-quality image or video.

This project contains a research and/or development component, as defined in the applicable law, and complies with Part 200 Uniform Requirements – 2 CFR 200.210(a) (14). nca/ncf