Noise elimination apparatus for eliminating noise components having a high density and a low density

Abstract

Color image data of which background is removed is converted into monochrome data and noise in the monochrome data is roughly discriminated in a first noise discrimination process. At the same time, the color and density of an attentional pixel, information area including a character or a picture, and others are checked, according to which operations the resulting first noise discrimination data is corrected in a second noise discrimination process. Substitute data is created from the image data of which background is removed according to the first noise discrimination data, and the substitute data substitutes for a pixel regarded as noise in the second noise discrimination process to obtain an image from which noise is eliminated. With the above arrangement, noise having a high density and noise having a size greater than the scanning window are eliminated without receiving influence from the surrounding noise.

Dependencies are inferred from the claim text.