Visually lossless Video compression

Obtaining experimental results using visually lossless compression in live video.

Research opportunity

Abstract: The image quality in conventional Shannon video is determined by the “hardware” or the bit rates in the transmission. The bit rates are determined by: the number of pixels on the screen, the bit per pixel color resolution, and the scanning rates. The images actually shown on the screen are irrelevant: a totally random noise image would have the same bit rate as a totally blank image. This is obviously not the way we humans perceive video images. In Autosophy video, in contrast, the bit rates are determined by the video content, such as motion and complexity within the screen images. The screen size, resolution and scanning rates are irrelevant. This provides very high “visually lossless” video compression, where the image quality depends on the video receiver. If the receiver cannot perceive the “difference”, between the original video images and the transmitted video images, then the image quality is “visually lossless”, according to the Autosophy information theory. This theoretical prediction must be verified by experiments using human test subjects. The purpose of this research is to determine the human perception limits, to select the resolution required in live television, before visual artifacts or distortions become visible.

Applications: This research is necessary to select the design parameters of practical High Definition Television (HDTV) systems. This may establish a transmission and quality standard to be acceptable to the public.

Keywords: Autosophy, Visually lossless video compression.

Available downloadable documents:

Publication 2006 – Satellite – Webpage htm

Publication 2006 – Internet video – Webpage htm

Demonstration – Autosophy video – Webpage htm