A shape-adaptive structural constraint
Y |
![](Neigh_af16_JPEG10-1.png) |
![](arrow_left.png) SA-DCT filtering |
![](Det_sdc_f16_lumchrom1_JPEG10.png) |
U |
![](Neigh_af16_JPEG10-2.png) |
![](arrow_left.png) SA-DCT filtering |
![](Det_sdc_f16_lumchrom2_JPEG10.png) |
V |
![](Neigh_af16_JPEG10-3.png) |
![](arrow_left.png) SA-DCT filtering |
![](Det_sdc_f16_lumchrom3_JPEG10.png) |
The shapes found on the luminance channel are used for the SA-DCT filtering of all three channels. Filtering performed on each channel independently from the others.
|