Incremental printing of symbolic information – Ink jet – Controller
Reexamination Certificate
1999-05-17
2002-06-11
Gordon, Raquel Yvette (Department: 2853)
Incremental printing of symbolic information
Ink jet
Controller
Reexamination Certificate
active
06402281
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an image processing method and device and an image formation device which converts input multilevel image data into multilevel image data with a smaller number of tones through a multilevel error diffusion process.
Conventionally, with image formation devices, such as printers, that use line heads such as line LED (light emitting diode) heads, line thermal heads, and line ink jet heads, dots of the same size are printed at a resolution inherent in the head used onto recording paper to form a bilevel image. With the line LED heads having a number of LEDs arranged in a horizontal line, the resolution corresponds to the spacing between each LED in the raster direction. Likewise, with the line thermal heads having a number of heating resistor elements arranged in a horizontal line, the resolution corresponds to the spacing between each heating element in the raster direction. With the line ink jet heads having a number of ink nozzles arranged in a horizontal line, the resolution corresponds to the spacing between each nozzle in the raster direction.
In these image formation devices, character images are simply reproduced as bilevel images with the specified resolution of their heads, while photographic images are reproduced through the use of halftoning such as ordered dither method or error diffusion method. In the halftoning process in this case, it is very difficult to make the maintenance of a high resolution and the reproduction of a high tone levels compatible with each other. In the ordered dither method in particular, resolution and tone levels conflict with each other.
A line head-based image formation device has recently made its advent which can represent one pixel with a few tone levels by using multi-valued image data and modulating the printing area within one pixel.
FIG. 15
shows a recording head having a number of recording elements arranged in a horizontal line and examples of dots recorded by the recording head
1
. For simplicity, in the example of
FIG. 15
, one pixel takes one of three values, including white. Arranging four or three such recording heads in parallel allows color images to be recorded by combinations of four colors of cyan (C), magenta (M), yellow (Y), and black (K) or three colors of C, M, and Y.
Such image formation devices that allow for the recording of multilevel image data perform various image processes, such as color conversion, undercolor removal (UCR) and/or gamma (&ggr;) correction, and then, in order to reproduce a specified number of tones inherent in the printer engine that actually performs image printing operations, perform multilevel halftoning using screen angle such as multilevel dithering or multilevel error diffusion process, thereby obtaining multilevel image data having several bits for each pixel. And the devices effect an improvement in image reproducibility by allowing one pixel to contain more information content.
In general, the ordered dither process is a light and fast process and can effect cost saving. In comparison with the dithering, on the other hand, the error diffusion method is complex but superior in image quality. The error diffusion method has been extensively used because its complexity has recently been solved by advances in LSI technology.
In many cases, line heads of ink jet line printers have variations in the volume and direction of ejected ink from each ink nozzle. In order to reduce such variations below a constant value, extremely high accuracy is required in manufacture, resulting in very high manufacturing cost. On practical side, it is therefore inevitable that the variations in the volume and direction of ejected ink occur from each ink nozzle. In the presence of such variations, dots formed by ink nozzles that tend to make large dots or adjacent dots the spacing of which is smaller than a standard value make the density of that portion high. On the other hand, dots formed by ink nozzles that tend to make small dots or adjacent dots the spacing of which is larger than a standard value make the density of that portion low or produce a white stripe in that portion. In either case, density nonuniformity occur, resulting in degradation of image quality.
In order to prevent such degradation of image quality, use has hitherto been made of an approach to, as in thinned-out printing in the checkered form, control the production of density nonuniformity appearing in the form of a stripe in particular by, for lines in the sub-scanning direction in which direction paper is fed, printing each line through the use of two or more ink nozzles instead of printing by a corresponding one of the ink nozzles to vary the dot size or the spacing between adjacent dots.
Such an approach allows density nonuniformity to be reduced to some degree but not enough. In particular, in the case of a type of line printer that allows one pixel to be represented in several tone levels by modulating the printing area within one pixel using multilevel image data, if a multilevel image is reproduced with dots of such intermediate size that adjacent dots barely come into contact with each other, there arises a problem that density nonuniformity in the form of a stripe becomes significantly noticeable. In view of the human visual system that is very sensitive to horizontal and vertical lines, it is quite possible that slight positional displacement is recognized as density nonuniformity in the form of a stripe.
There is another approach to control density nonuniformity by finding characteristic values for each ink nozzle in advance through test printing, storing corrections for all the ink nozzles in memory, and, at the time of printing, correcting the printing characteristics of each ink nozzle. The application of this approach to a print head, such as a line head in which a large number of ink nozzles are arranged, requires a very large capacity memory to store the corrections. In addition, a driver LSI for correcting the printing characteristics is required to have significant control capabilities. Thus, the approach is difficult to implement.
BRIEF SUMMARY OF THE INVENTION
It is an object of the present invention to provide an image processing method which, in a range of low tone levels in which stripes and density nonuniformity are relatively difficult to be noticed, maintains the substantial resolution by reproducing dots of the same size for image data of the same level and, in a range from medium to high tone levels, reduces density nonuniformity by reproducing dots of multiple sizes for image data of the same level, and moreover is easy to implement.
According to a first aspect of the present invention, there is provided an image processing method for converting input multilevel image data in which each pixel consists of M bits of data into output multilevel image data in which each pixel consists of N (M>N≧1) bits of data so that the output multilevel image data has a smaller number of tone levels than the input multilevel image data through a multilevel error diffusion process and producing dot patterns according to the tone levels of the output multilevel image data, wherein the multilevel error diffusion processing is such that, when the input multilevel image data is in a range from medium to high tone levels, the types of dot patterns produced by the output multilevel image data are larger in number than those produced when the input multilevel image data is in a range of low tone levels.
It is another object of the present invention to provide an image processing device which, in a range of low tone levels in which stripes and density nonuniformity are relatively difficult to be noticed, maintains the substantial resolution by reproducing dots of the same size for image data of the same level and, in a range from medium to high tone levels, reduces density nonuniformity by reproducing dots of multiple sizes for image data of the same level, and moreover is easy to implement.
According to a second aspect of the present invention, there is provided
Nakahara Nobuhiko
Umezawa Hiroki
Frishauf Holtz Goodman & Chick P.C.
Gordon Raquel Yvette
Toshiba Tec Kabushiki Kaisha
LandOfFree
Image processing method and device and image formation device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method and device and image formation device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method and device and image formation device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2955862