The Ikena Copyright system produces a collection of hashes from existing video content. These hashes are then run against new video to attempt to find matches. Unlike other fingerprinting technologies, Ikena Copyright generates hashes based on motion and not color or audio. A problem facing video hashers is that it is not difficult to obscure video attributes such as color, size, audio or resolution. Most hashing software is easily fooled by the traditional bootleg video which is made by recording directly off a TV or Movie screen. Ikena Copyright, MotionDSP claims, can detect such bootlegs because it is much more challenging to alter motion. Ikena Copyright’s major competitor is Audible Magic, which filters video for Google’s YouTube.

MotionDSP’s video enhancement product takes low quality (camera phone quality) video and renders it clearer and less pixilated. Its software compares video frames which are close together and deduces which pixels are missing or altered, then fixes them. The end product is not much larger than the original file, conserving much needed space. This process can be done in real time, meaning that a minute of video can be enhanced in one minute. It works best on videos that are reasonably stable and seems to work best on faces (see example).

Above: MotionDSP Screenshot -- #1