Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
1999-12-17
2004-03-09
Grant, II, Jerome (Department: 2626)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S003150
Reexamination Certificate
active
06704123
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to image reproduction and, more specifically, to a method for applying tonal correction to a binary halftone image.
BACKGROUND OF THE INVENTION
In order to reproduce a realistic image of a scene, the reproduced image must have a believable tonal representation of the original scene. Tonal representation refers to the relative values of tone within an image (hereinafter referred to as “relative tonal values”), which help provide important information about the image, such as surface characteristics. Typically, these relative tonal values are measured at fixed spatial coordinates on a rectangular grid where each sample point is referred to as a picture element, commonly known as a pixel.
Image-reproduction systems fall into two categories: (1) systems capable of rendering continuous tone (contone systems); and (2) systems with limited tone reproduction capacity (binary systems). In contone systems, pixels can take on a wide range of tonal values. Examples of contone systems include photography and television. In contrast, pixels in binary systems can only take on one of two tonal values. These values correspond to either an “on” or an “off” state. For example, in printing devices, such as offset or flexo presses, each pixel on a printed medium can either be covered (“on”) or not covered (“off”) by ink. For color printing, each pixel can either be “on” or “off” for each layer of a different color ink. Such layers are commonly referred to as color separations.
Since images to be reproduced by such binary systems are often initially contone images, these binary systems use a process called “halftoning” to achieve a wider range of tonal values (gray or color levels). Thus, the binary systems achieve a believable tonal representation of the original contone images. In general, halftoning is a process for approximating continuous-tone values using a pattern of pixels. Various halftone processes are well known in the art, such as error diffusion and dithering.
Basically, the halftoning process defines a region of pixels as a halftone cell, sometimes referred to simply as a “cell”. The halftone cell contains a specific, repeatable pattern referred to as a “halftone pattern” that is typically beyond the resolution of the eye. For each “halftone pattern,” the pixels are numbered in a specific order. The specific order determines the sequence for turning pixels “on” in the halftone cell. The number of pixels in the cell determines its tonal range. For example, a halftone cell with a two-by-two matrix of pixels may represent five different intensities of a color (also commonly referred to as shades or tonal values): white; black; and three intermediate levels of gray. The color white is represented by all four pixels in a white or “off” state. Conversely, black is represented by all four pixels in a black or “on” state. The intermediate levels of gray are represented by one, two or three pixels in the “on” state and the remaining pixels in the “off” state.
A tonal value for a specific location “seen” by the eye (hereinafter referred to as a “visible tone”) is approximated using the halftone pattern in the neighborhood of the location, i.e. approximately the percentage of “on” pixels in the neighborhood of the specific location. The visible tonal value is only approximately, and not exactly proportional to the percentage of “on” pixels, due to a variety of reasons related to the printing environment. Some of such reasons are that printing devices do not produce ink spots with exactly the same size and shape as a pixel; light reflection and diffusion in the selected paper affect the tone; hue and density of the ink may differ between printing devices. Therefore, in order to achieve the desired tonal representation for a specific printing device, a calibration curve is applied before printing. The calibration curve is specific to each individual printing system and is derived experimentally using methods well known in the art. The calibration curve describes the required percentage of “on” halftone pixels for generating each tonal value for a specific printing environment. Some calibration curves describe this relationship for each color separation. However, other calibration curves describe the required number of “on” halftone pixels for each color separation based on the overall desired tonal value taking into consideration the interaction between the color separations. The process of applying a specific calibration curve to a halftone image is referred to as “tonal calibration” or “tonal correction”. A halftone image that has a percentage of “on” pixels proportional to the original image is referred to as an “uncalibrated” image because the desired tonal value of the original image is achieved without modifying the tonal values of the pixels.
Preferably, when a contone image is received, a binary image reproduction system performs tonal correction utilizing a calibration curve associated with the target printer before performing or completing the halftone process. However, in certain situations, the target printing device may acquire a digital halftone image or a halftone film that is subsequently converted to a digital halftone image by a copydot scanner. In both these situations, because halftoning is completed before the target printing device and other printing conditions are known, the halftoning is performed either without tonal calibration or using an incorrect calibration curve. This results in artifacts in the printed image.
One prior attempt for applying tonal correction to a binary halftone image involves a three-step process: (1) descreening the original halftone image to create a contone image; (2) adjusting the tone of the contone image; and (3) screening the contone image to produce a new binary halftone image with the adjusted tone.
The descreening step applies a low pass filter to the binary original halftone image to produce the contone image. The low pass filter removes the original halftone structure from the original halftone image. A disadvantage of this step is that a filter strong enough to remove the halftone structure also removes some image detail. The adjusting step replaces each pixel in the contone image with a pixel having a corrected tone. The pixels having the corrected tone may be derived from an arbitrary function of the tone of the original pixel or may be derived from a more complex function of the tone taken from multiple color separations. The final screening step creates the new binary halftone image from the contone image. This step takes as input the contone image, screen frequency, screen angle, and dot shape. The screen frequency and angle may be extracted from the original binary halftone image with good success using Fourier techniques. A disadvantage of the step is that the dot shape can not usually be determined very reliably and choosing the wrong dot shape may cause image artifacts, such as tone jumps.
Another prior attempt for applying tonal correction to a binary halftone image involves using a low-pass filter that retains the halftone structure. Thus, the low-pass filter only softens the edges of the features in the image. Then, after descreening with the low-pass filter, a threshold is applied to the contone image to approximately reconstruct the original binary halftone image. If a point in the image is whiter than the threshold, the point is set to pure white. If a point in the image is blacker than the threshold, the point is set to pure black. In order to reconstruct the binary halftone image accurately in the highlights and shadows, the threshold must change as a function of the tonal value at each point. Adjusting the threshold away from the nominal position can be used to perform tonal correction on the image. This approach has two disadvantages. First, the filtering step removes a small amount of spot shape detail from the original image. Second, because the tonal change does not account for dot shape and edge roughness, errors are introduced during the tonal correct
Av-Shalom Amit
Bielak Richard Roman
Creo Inc.
Grant II Jerome
Oyen Wiggs Green & Mutala
LandOfFree
Method for applying tonal correction to a binary halftone image does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for applying tonal correction to a binary halftone image, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for applying tonal correction to a binary halftone image will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3210796