Input gray-scale bit map information is inputted to an image forming position extracting unit for extracting image forming position information. An input computer-formed image is inputted to an image forming position extracting unit for generating an image forming position. At an image forming position overlapping unit, two image forming positions are overlapped with each other, and the composed image forming position is sent to a pixel value calculating unit and an image forming position determining unit. The image value calculating unit measures a width of the composed image forming position and converts the measured value into a pixel value. The image forming position determining unit calculates a barycenter of the composed image forming position. The pixel value and the barycenter are sent as image information. Encoding without distortion can be executed by enhancing the precision of the barycenter.
An image input section outputs the size of an image to a division-number setting section and outputs the image to an edge enhancing section, which enhances the edges of the image. An edge extracting section extracts the edges, an edge evaluating section checks whether each pixel belongs to an edge, and an edge counting section outputs the frequencies of the edges. A DFT section applies a Fourier transform to the edge frequencies which are output as power spectra, and a peak extracting section outputs the spatial frequencies at the peaks of the power spectra. A division-number setting section determines the number of divisions from the spatial frequencies. An image dividing section divides the image into the determined number of blocks and a colored-image output section assigns a particular pixel value to all pixels in each of the blocks.
This invention has as its object to prevent image quality deterioration of vector and raster image parts by allowing adjustment of the toner amount for respective objects when single image data includes both vector and raster image parts. To accomplish this, image data is separated into vector and raster image parts, a vector or raster image attribute is assigned to each pixel, and the toner amount is adjusted for each pixel of each object.
