Pasadena, CA (PRWEB) June 30, 2012
Cognitech, Inc. has announced the receipt of the second related patent award from the United States Patent and Trademark Office, for the invention of the “Video Demultiplexing Based on Meaningful Modes Extraction.”
The “Video Demultiplexing Based on Meaningful Modes Extraction” invention relates generally to automatic video analysis and sorting of people, vehicles, and objects in video, and in particular to video demultiplexing. Video demultiplexing is needed for viewing and forensic processing of CCTV videos, which combine sequential recordings from multiple surveillance cameras. Since there is no uniform format in which such CCTV systems multiplex (i.e. combine) video streams from multiple video cameras, each video recording requires the specific corresponding demultiplexer hardware for video viewing and for video capturing for subsequent forensic processing. Cognitech's scientists were the first to publish a mathematical breakthrough solution for the demultiplexing problem, in the 2001 Proceedings of SPIE, which showed a universal software based demultiplexing method, thus eliminating a need for the format-specific expensive hardware. Cognitech's [Video Investigator® was historically the first forensic video software system to contain the software-based universal video demultiplexing module. Since then, many other competing forensic video software products have attempted to replicate the capabilities of the original Cognitech demultiplexing algorithm. Even over a decade later, they lack the accuracy (Cognitech demultiplexing is flexible to the dynamic scene changes, such as the number of people walking), the speed (Cognitech demultiplexing can be done in real-time) and the versatility (such as the ability to demultiplex video streams with common CCTV pan/tilt/zoom) of Cognitech's demultiplexing.
“Video Demultiplexing Based on Meaningful Modes Extraction” is in part a fully automatic method for grouping video frames accordingly to their source cameras or source internet video streams. Accordingly to the awarded Cognitech US Patent specification, “Video frames are grouped by determining if the video frames share meaningful modes for values representative of features of the video frames”. This is the next big step in the evolution of the demultiplexing algorithms. First, it does not require any laborious manual input from the end-user, that all competing systems necessarily do. Such inputs include the individual camera views samples, the number of cameras in the multiplexed video, the accuracy prone thresholds that separate camera views, etc. Second, Cognitech’s patented algorithm is so general that anyone attempting to implement an automatic or threshold free demultiplexing method, will necessarily have to infringe on Cognitech's patent. Third, Cognitech's patented algorithm extends the definition of video demultiplexing to a more general computational ability to sort video accordingly to their content, such as specific people, vehicles, and objects that are present in the video. Forensic investigators have been long seeking general video sorting and searching tools to assist with their investigations. Cognitech's Video Demultiplexing Based on Meaningful Modes Extraction has opened a wide door for such applications.
Cognitech’s CEO, Dr. Lenny Rudin said, in regards to this invention," It is not an every day event, to be awarded a coveted US Patent. This comes as a result of a prolonged and concentrated effort by Cognitech’s R&D team to push the envelopes of science and technology, which generally benefits the field of justice and in this specific instance it benefits the entire field of forensic video analysis. The end-users of Cognitech's Forensic Video Tri-Suite software are the immediate beneficiaries of Cognitech's latest technological breakthrough, since the current release of the Cognitech® [VideoActive® software, the real-time forensic acquisition and processing software that is unique to Cognitech, includes the fully automatic module implementing this US Patented Video Demultiplexing Based on Meaningful Modes Extraction.
About Cognitech, Inc.:
In 1988, Cognitech, Inc. was the first company that designed and developed the unique Video Investigator® hardware and software products for professional forensic CCTV video applications. Cognitech has demonstrably by far the most extensive US Patents portfolio in the image/video processing and analysis, among all of the companies in the forensic video field worldwide. The recent scientific breakthroughs of Cognitech’s research team, and a non-stopping stream of USPTO Awards, including this very recent US Patent for Video Demultiplexing Based on Meaningful Modes Extraction, proves that twenty four years later, Cognitech continues to be the non-rivaled technological innovation leader in the forensic image/video processing field.