An example of JPEG masking that shows the contrast masking effect in action. Bovik applied insights from visual psychology and neuroscience to design tools that allow broadcasters and streaming video sites to compress and distribute video so its distortion is minimized. In the above image, the distortion is the same everywhere, but is less visible in busy regions (in the grass and flowers or the girl's hair) and more visible in smooth regions (on the girl's face). This does not occur because the grass hides the distortion, but instead, is a by-product of a visual process that renders the distortion invisible or nearly so. SSIM is designed to predict these 'content-dependent' effects.

Credit: The University of Texas at Austin

In the following image, all distorted images have roughly the same mean squared error values with respect to the original image, but very different quality. SSIM gives a much better indication of image quality.

Credit: Zhou Wang, Alan C. Bovik, Hamid R. Sheikh and Eero P. Simoncelli

Alan Bovik holds the Ernest J. Cockrell Endowed Chair in Engineering at The University of Texas at Austin, where he is a Professor in the Department of Electrical and Computer Engineering and The Institute for Neurosciences, and Director of the Laboratory for Image and Video Engineering (LIVE).

Credit: The University of Texas at Austin