Euclidvision

From Wikipedia, the free encyclopedia

EuclidVision is a proprietary video compression codec developed by Euclid Discoveries. Euclid Discoveries claims to build upon existing MPEG-4, adding extensive object-based compression, such as a refined method for compressing human faces, manipulating object planes, or 3D object modeling. The video compression achieves a 460% improvement over MPEG-4, and 600% improvement over DVD, MPEG-2 compression. The company claims that today's 700MB MPEG-4 movie could be compressed to 50MB with EuclidVision. It can compress standard video down to a phenomenal 4kB/s with what the company claims results in "...a clear, recognizable video." The technology was tested on a simulated-constrained network with 9kB/s of bandwidth, similar to the requirements of a mobile phone. The current technology is designed for video conferencing over twisted pair copper telephone wire. The company has filed 25 patents that cover every minute aspect of the technology so that no competitor can create a similar video compression method. ^[1]The ability of EuclidVision to survive in the video compression market will determine the future usage of the video codec.

EuclidVision’s object based compression “identifies individual objects shown in a video, then calculates the optimum level of compression for each of them.”^[2] Further, the algorithm remembers the objects and stores them in memory a single time so that even if an object appears multiple times, it is saved once. Euclid Discoveries was recently granted Patent #7,158,680 ^[3] for “apparatus and method for processing video data.” This technology could greatly enhance broadband delivery over cable networks and the Internet as higher definition standards are being released.