| 摘要 |
<p>Image artefacts, such as block noise, are identified following decompression of an image signal and subdivision into regions (<B>310</B>), each containing a number of pixels (<B>322</B>). The subdivision may be controlled by user selected parameters. Each region is investigated to identify any artefacts, such investigation being executed according to selected parameters. Any region determined to contain artefacts is smoothed by a filter (e.g. using data values of surrounding pixels) to remove picture element discontinuity. The smoothed regions are reinserted into the decompressed image, ensuring a high quality output.</p> |
