| 摘要 |
A method of generating video data for transmission to a user particularly fo r use in a video surveillance system. The method comprises generating a first representation of a first image and one or more further representations of t he first image are then generated, said further representation(s) being predict ed from a previously generated representation of the first image. In response t o a request for a subsequent image, a first representation of said subsequent image being predicted from a representation of the first image. Then one or more further representations of said subsequent image are generated, said further representations of said subsequent image being predicted from a previously generated representation of said subsequent image. Thus, the same source data for a first image is fed into the encoder, so producing a progressive still image at the decoder. When a different image is required, the encoder substitutes, as the input to the encoder, the source picture fro m the requested point in time. This source picture is encoded predictively fro m the original image.
|
