| 摘要 |
A scheme for digitally encoding continuous-tone information from an original pictorial image includes processing electrical signals via separate networks (20,30) to generate (a) high-resolution detail data of "black" and "white" pels, and (b) low-resolution gray-scale or background data approximating the grayness of two-dimensional blocks (104) of pels (102). In the gray-scale network (30), the electrical signals are first compared (31) to different position-dependent threshold values to generate a binary signal (108) of black and white pels which represents gray-scale. Then, any of the pels (102) determined "black" in the detail network (20) are subtracted (36) from the gray-scale binary signal to generate a modified gray-scale binary signal (110). The black pels in the modified gray-scale binary signal for each block (104) are counted (42), the count approximating the background gray level of each particular block (104) of pels (102). Subtracting the detail "black" pels generates a lighter gray level for each block (104) of pictorial area in which detail "black" pels are found. Each block count is further compressed and encoded to represent the background gray-scale information. |
