A Fast Inverse Half toning algorithm is used for error diffused halftones. Halftones and other binary images are difficult to process cause several degradations. Jarvis’s halftoning operator is used to obtain sharper halftone image. Color image of 24-bit BMP formats splits image into 8- bit R-G-B planes. Degradation is greatly reduced if the halftone is inverse halftone (converted to gray scale) before scaling, or other processing. At each pixel, the proposed algorithm applies a separable 7 x 7 FIR filter. The filter passes in vertical and horizontal direction separately and computes vertical and horizontal edge strengths from local gradients. It requires entirely local operations and storage of 7 rows. We compare our algorithm with previously reported approaches to show that it delivers comparable subjective quality at a fraction of the computation and memory usage.