Video and moving image compression

Refers to techniques used to eliminate redundant information of moving image files in order to reduce its size, thereby facilitating its transmission. Like other forms of compression, video compression can be either lossless or lossy. However, as lossy compression produces considerably smaller files that maintain an acceptable quality, lossless compression is rarely used for video.

Lossy video compression can be divided into intraframe and interframe compression, and deals with spatial and temporal redundancy.

In intraframe compression, spatial redundancy is reduced by registering differences in the pixels, or blocks of pixels, within a single frame. It works in a similar way to image compression, by ‘averaging out’ small differences in colour that are imperceptible to the human eye.

In interframe compression, temporal redundancy is reduced by registering differences between frames, rather than registering each entire frame, which would often involve duplicating a large number of the pixels. In this way, scenes with little motion can be compressed more effectively than fast scenes with detailed visual aspects.

If the compressed video is edited, interframe compression can cause problems as if the original frame in a scene is deleted, subsequent frames will not play properly as they cannot be reconstructed. This is not an issue with intraframe compression.

Video can also be compressed by the use of variable bitrate (VBR) encoding (see sound encoding), where a higher bitrate is used in fast moving scenes, and a lower bitrate is used for scenes in slow motion.

Related methods include: Image enhancement, Sound compression and Sound encoding.

Syndicate content