| 摘要 |
For reducing the false contour effect the gravity centre coding is utilized. However, a GCC with a few number of levels introduces noise in a picture. Therefore, each picture is divided into segments and such GCC is performed only in regions, where the false contour effect is critical. This is the case for skin areas of a picture. Such areas are detected by the colour analysis and texture analysis. If this analysis leads to a skin area, a specific false contour mode for data processing is used. Otherwise, a standard mode is employed.
|
