Packet Video Transmission on the Information Superhighway Using Image Content-Dependent Autosophy Video Compression

Designing the Information Superhighway with Autosophy video compression

Authors: Klaus Holtz,

Published / Presented at: IS&T’s 48th Annual Conference. Imaging on the Information Superhighway. May 7 – 11, 1995, Omni Shoreham Hotel, Washington, D.C.

Level: Expert Internet – Information Superhighway

Abstract: Transmitting live video on the packet switching Information Superhighway using conventional Shannon-based techniques will be very difficult and expensive. The MPEG-2 image compression standard can reduce transmission cost but only by greatly reducing image quality. Autosophy television is an alternative based on a new mathematical theory of “Learning” and a new Information Theory in which “Information” is defined by image content rather than by hardware as in conventional television. Transmission bandwidth becomes virtually independent of hardware factors such as screen size, resolution or frame rates; depending instead only on “Novelty” (that which has not been seen before) and “Movement” (change) within images. The new television provides great lossless image compression, nearly unbreakable encryption and a transmission code especially suitable for packet switching networks such as ATM or the Information Superhighway. Autosophy image compression – based on “Parallel” Omni Dimensional Networks – has been described in many publications and tested in software simulation experiments, which show very promising results. Alternatively, this paper will explore image compression based on “Serial” Omni Dimensional Networks, which are already used commercially for data compression in V.42bis modems.

Available downloadable documents:

Publication Top Page -- Adobe pdf

Published document -- Adobe pdf