摘要 |
The simulation of film grain in a video image occurs by first creating a block (i.e., a matrix array) of transformed coefficients for a set of cut frequencies f<SUB>HL</SUB>, f<SUB>VL</SUB>, f<SUB>HH </SUB>and f<SUB>VH </SUB>associated with a desired grain pattern. (The cut frequencies f<SUB>HL</SUB>, f<SUB>VL</SUB>, f<SUB>HH </SUB>and f<SUB>VH </SUB>represent cut-off frequencies, in two dimensions, of a filter that characterizes the desired film grain pattern). The block of transformed coefficients undergoes an inverse transform to yield a bit-accurate film grain sample and the bit accurate sample undergoes scaling to enable blending with a video signal to simulate film grain in the signal. |