| 摘要 |
A method and apparatus of producing gray level images comprises controlling a gray level printhead such that the printhead forms dots of variable dot sizes on a recording medium at plural pixel locations, the pixel locations being grouped into cells having cell gray levels, wherein the dots of a cell are determined such that for each increase in cell gray level, a dot at at least one of said pixels in the cell forms a larger dot-size. Cell growth patterns for full dot, partial dot and mixed dot are described. In another embodiment, statistical information of dot regions from a digitized image signal are collected and a dot type to render that cell dot region in accordance with the statistical information is provided for. In a further embodiment, a gray level printer forms dots along lines in accordance with at least a first template corresponding to increasing cell gray levels until a first line structure is stable and the gray level printer forms dots along lines in accordance with at least a second template corresponding to further increasing cell gray levels. In still another embodiment, pixels in each cell are classified into one of at least two classes and an exposure level for printing said pixels is modified according to the classification. In another embodiment, scanned pixel values are compared to a set of thresholding values that are defined by a 4-quadrant tone reproduction analysis using a dot template, selected bit depth and dot type structure. A gray level halftoned representation of the image is generated using the thresholding values. In another embodiment, the intensity values of some pixels are modified by introducing errors to intensity values of pixels at preferred locations such that pixels along screen lines have a greater likelihood to be printed with stable dots than pixels not at such preferred locations. Errors, created by modifying the intensity values of pixels at such preferred locations, are diffused to pixels neighboring the pixels at said preferred locations. In still another embodiment, a region of an original image is analyzed to determine if it contains either a text region or a halftone region, and contour suppression is performed on every pixel of said region. In accordance with another embodiment, a region of an original image that contains either a text region or a halftone region is printed by a gray level printer with an output screen frequency higher in screen frequency than the screen frequency of the original image. In still another embodiment of the invention, a descreening filter includes means for transforming the digitized image into a frequency domain, means for reducing the strength of a specific frequency content from the digitized image, and means for transforming the digitized image back into the spatial image domain and feedback means between the spatial domain and the frequency domain to iteratively reduce to a minimum the specific frequency in the digitized image. <IMAGE> |
