Video codec programs are currently capable of reducing the file size (bandwidth) of video content by a factor of about 100:1. A recently developed algorithm utilizing the Haar wavelet transform (WT) and interfacing with a conventional codec program can reduce the size and bandwidth requirements of a video file by an additional order of magnitude, by a factor of 10, without loss of video quality. In other words, when a conventional codec alone decreases a video file to 1/100 of its original uncompressed size, then the combination of codec and Haar WT algorithm decreases the size of the video file to roughly 1/1000 of its original size.
Industry commentators in recent years have reported a growing crisis in telecommunication network capacities. Global bandwidth demand has increased exponentially with the proliferation of new services and products, such as mobile broadband applications for laptops and smart phones, internet protocol television (IPTV), video on demand (VOD), video downloading and streaming video. Video will soon replace voice calls as the preferred real-time communications medium Live Sports.
The so-called bandwidth crisis is two-fold. Telecommunication infrastructures were planned and built for a world that did not contemplate ubiquitous video files. Nevertheless, service providers and manufacturers continue to introduce new products and subscription packages to increase demand and market share, leading to bandwidth capacity shortages.
Bandwidth providers could invest capital to increase capacity. Costs would be passed along to unhappy customers. Or, they could raise user prices to decrease demand, also making customers unhappy.
A better solution for service providers, device manufacturers and end-users is to decrease video file size, thereby freeing up capacity, decreasing data storage and transmission costs, allowing continued market and revenue growth, and avoiding service restrictions.
It was not known until recently that the Haar wavelet transform (WT) could be effectively applied in the field of video compression and decompression. A wavelet is a mathematical function localized in space. As such, wavelet transforms are useful for analyzing physical phenomena having discontinuities and spikes, such as occur in audio signals and images. A WT can extract essential information from data and thereby compress it. The Haar WT is one of the simpler wavelet transforms and is, therefore, relatively easy to utilize.
Human beings process sensory information in their brains in a manner somewhat similar to wavelet transforms, that is, they extract important information and discard data that is irrelevant from the standpoint of human perception. Accordingly, an algorithm based on the Haar WT can be designed to mimic the data processing of human brains. It compresses video data by extracting information vital to human perception and discarding less relevant data.
A new video compression-decompression software module interfaces directly with standard codec (e.g., MPEG-4, H-264, VC-1) to reduce video data file size by a factor of about 10:1 beyond the compression achieved using a codec alone. The new software achieves a corresponding reduction in bitrate. Video quality is preserved. Conventional codec programs remain unchanged. The special video compression (and decompression) algorithm based on the Haar WT is designed to be virtually plug-and-play, requiring only minor adaptation to interface with a particular service provider’s codec.
Thus, an algorithm using the Haar WT can immediately cut the bandwidth demands of video content on cable, satellite and wireless transmission networks by 90 percent across the board, most importantly in the mobile broadband sector. Further, based on current prices of storage and transmission, video content providers can cut their storage and transmission costs by 60 to 80 percent.