JP2002101307A - Method and system for generating half tone output image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for enhancing a binary and multi-bit half-tone technology. SOLUTION: A half-tone output image is generated from a source image by a prescribed process. The output image has plural half tone tables (HT) each provided with half tone cells (HC), and each HC relates to an array of a pixel value decided by a threshold value HT value having threshold values HC relating to an array of pixel positions. The source image is expressed as an array of intensity values. A cell growing pattern provided with the array of the pixel positions in the threshold value HC is specified with respect to each threshold value HC in the threshold value HT. The characteristic of the process gives effect on the array of the threshold values with respect to each threshold value HC in the threshold value HT. A full growing pattern provided with the array of the pixel positions in the threshold value HT is also specified with respect to the threshold value HC and the characteristic of the process gives effect also onto the specification of the full growing pattern. Each intensity relates to the HT of an output image. The pixel value of each HC in each HT is set through the comparison between the related intensity and the threshold values in the threshold value HT.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a digital copying machine,
It relates to the reproduction of images, color or monochrome halftones by reproduction devices that can render more than one tone value, such as electrophotographic printers, ink jet printers, scanners and other devices.

[0002]

BACKGROUND OF THE INVENTION The problem of converting a continuous tone image input into an encoded representation that is compatible with mass reproduction of the image has a long history and is well known in the art. I have. Indeed, Stoffel entitled "Overview of Electronic Technology for Image Reproduction" in IEEE Journal, Vol. 29, No. 12, December 1981 (St.
According to the literature by Offel) et al., the history of mass reproduction techniques for images dates back to the 8th century and includes historical technologies such as relief printing (letterpress), intaglio (photo intaglio printing) and lithography. In each case, the most efficient mode for their operation is
Employing halftone technology, the image was reproduced here by employing a microstructure consisting of regions with only two consistently reproducible tone values corresponding to the presence or absence of ink. As technological advances improve the spatial resolution of binary output devices, the variable darkness tone will increase the size and / or distance between "halftone dots", separate units within the microstructure. By modulating it could be represented in such a device.

[0003] Halftoning is therefore a process in which a continuous tone color or monochrome image is simply approximated by a reproduction process that allows for discrete tones and discontinuous tones. A continuous tone image is an image that is "natural" as far as humans can observe it, and includes apparently continuous intensity levels rather than a set of discrete intensity levels. An example of a continuous tone image is a photograph.
The continuous tone image may be reproduced using a discontinuous tone printer that approximates the continuous tone image composed of individual pixels at the printer resolution. In this case, a pixel is defined as the smallest addressable unit of a printer or other display device. Thus, if the resolution of the printer is greater than the resolution of the human eye, digital halftoning is a method that gives the human eye the illusion that the printer can create continuous tone.

[0004] The process of digital halftone reproduction of an image involves arranging the pixels having the output of the printer device into a predetermined pattern within the cells of a halftone screen used to further divide the source image into distinct regions. Is executed by The halftone screen is
As a result, it has a separate area, a halftone cell, that can be repeated in the horizontal and vertical directions, or at any angle, until it covers the entire image. As is well known in the art, two characteristics of a halftone screen are screen frequency and screen angle. The reciprocal of the absolute dimension of the halftone cell, that is, the number of cells per unit area of the halftone screen,
Also known as the screen frequency. On the other hand, the angle at which the cell repeats with respect to the reference horizontal and vertical lines is known as the screen angle.

[0005] An intensity level may be generated for each halftone cell corresponding to the intensity of the tone for the region of the source image represented by the halftone. For a monochrome (ie, black and white) reproduction of a halftone, a signal intensity value may be generated for each halftone cell corresponding to an intensity level for each region in the source image. Color reproduction can be achieved in a similar manner using multiple intensity values for each halftone cell. Here, an intensity value may be associated with each tone used in the color reproduction process.

[0006] Intensity values are typically generated by periodically sampling a continuous tone image using an optical scanner. Each intensity value represents an image intensity in an adjacent area surrounding the location in the continuous tone image from which the sample of the intensity value is obtained. Typically, each intensity value can be quantized into a plurality of levels known as intensity or grayscale levels. Quantization allows each intensity value to be represented by a digital value and processed by a digital circuitry into a digital halftone image. For example, if the intensity values are quantized to 256 levels (ie, 256 levels of gray scale), the intensity values may be represented by an 8-bit digital word.

[0007] Each halftone cell can be sequentially associated with a distinct plurality of pixel locations in the output image. The pixel location associated with each halftone cell is typically N
It has a rectangular array of × M pixels. A threshold table may be generated for the output image reproduction device having a plurality of entries each corresponding to a pixel location associated with a halftone cell. Given an associated intensity value, an entry in the threshold table determines how many pixels at a time are shaded from half-tone pixels to N × M shaded pixels in a halftone cell.

If the screen cells are small compared to the resolution of the observer's eye at the normal spacing for reading, that eye will integrate the individual pixels in the halftone screen and not individual pixels, You will see an image of a given color and density in halftone. Thus, if there are enough halftone cells and, more importantly, the pixels associated with the halftone cells in the halftone screen, the observer will actually be able to quantize to a separate halftone with multiple pixel locations. Even if you are watching a digitized image, you will think that you are watching a continuous tone image.

Each pixel has a limited number of output tone levels. For the production of monochrome images, these output tone levels are also known as gray levels. In a rendering device that employs binary halftoning techniques, each pixel location can produce only two gray levels, corresponding to black and white, shaded or unshaded, on or off. For example, if the size of a halftone cell is 4 × 4, there are 16 pixel positions in the halftone cell. If no pixels were turned on, the cell would appear white. If 16 pixels are on, the cell looks black. If only eight pixels are on, the cell appears to be a certain degree of gray (eight pixels white and eight pixels black). The degree of this gray appearing in the halftone cells varies depending on the arrangement of the black and white pixels with respect to each other. The variability of the degree of gray represented in a halftone cell can be represented by a spot function.

The separate halftone cells form the nucleus of the spot function. The spot function may consist of N × M distinct pixel locations of the halftone cell. In this case, as described above, a pixel is the smallest addressable unit of a display device. The spots generated by the spot function are the basis of what the eye sees when viewing the halftone image in close proximity (ie, individual spots / dots). In the basic case, the spot function is on (black) or off (white) in a 1 × 1 matrix, ie, binary halftoning
Or as small as individual pixels set to multiple different levels in multi-bit halftoning. More generally, the spot function may have a plurality of individual pixels, in which case
Each pixel in the spot function corresponds to a pixel in an N × M halftone cell. If the spot function has multiple pixels, the spot function can be used to control the degree or manner in which a pixel changes value in a halftone dot.

In operation, different intensity values that can be represented in a halftone cell may be generated using a threshold table. The threshold table has a number of elements corresponding to the number of pixels in the halftone cell. here,
Each element of the threshold value table corresponds to a pixel in the halftone cell. A threshold table with multiple halftone cells, on the other hand, may have more total pixels than a threshold table with a single halftone cell, while retaining the halftone cell structure in the threshold table. This provides greater flexibility in defining the spot function and, as a result, allows a larger range of intensity values to be represented using a threshold table while retaining the reference screen frequency.

Once the structure of the threshold table has been determined, the elements in the threshold table can be defined. The elements of the threshold table have thresholds that can be compared to the intensity values generated for the image. The result of this comparison determines the output value of the pixel associated with the threshold. For example, a continuous tone monochrome image may be quantized and represented as a matrix of intensity values. In this case, an arbitrary intensity value is associated with a halftone cell or other grouped pixels such as a halftone table having a plurality of adjacent halftone cells (used with the threshold table described above). On the other hand, each value may correspond to a pixel in a halftone screen.

In a binary halftoning system, any intensity value generated for a pixel (or a plurality of pixels or cells) in a halftone screen is compared to a threshold in a corresponding threshold table. Each pixel in the halftone cell corresponds to a threshold in the threshold table. If the threshold value of the pixel of the halftone cell is less than the intensity value associated with the halftone cell, the pixel is blackened, otherwise the pixel is whitened. As a result, the intensity values are mapped to an area consisting of an arrangement of black and white pixels whose total intensity is proportional to the magnitude of the intensity value.

The placement of thresholds in a threshold table is well known in the art and is commonly known as dithering. Some widely used types of dithering have dense dot dithering, where the first pixel grown is near the center of the matrix, and subsequent pixels form ) Assigned to a spiral pattern that extends in all directions towards the outermost pixel.
Another widely used type of dithering is dispersed dot dithering. In this case, the pixels of the halftone cell grow in a regular manner and are evenly distributed within the cell. In essence, through dithering, the threshold is arranged to ensure that the resulting halftone pixel accurately reflects the intensity of the input intensity value. For a review of dithering in systems of monotonicity, see Ulichney's Digital Halftoning, pages 71-171 (MIT Press, 1987).
Year). The threshold table comparison process is repeated for each intensity value sampled from the original continuous tone image. As a result, the entire image is spatially modulated into a halftone image consisting of a tiled arrangement of halftone cells each representing an intensity value sample.

Also, as known in the art,
The binary halftoning process described above repeats the binary halftoning process for each primary color, ie, red, blue and green, or cyan, magenta and yellow, followed by color halftoning with proper alignment. Overlaying the image is useful when halftoning a color image.

The same principles used for binary halftoning techniques can also be used for multilevel halftoning techniques. As the name implies, multilevel halftoning replaces each black or white pixel in a binary halftone cell with a pixel having a grayscale or tone value selected from a series of useful grayscale levels. . Many display devices allow for the display of multi-level pixels. That is, multi-level halftoning takes advantage of this performance. For example, a thermal printer can print a dot size corresponding to each pixel intensity level.

[0017] Typically, in multi-level halftoning, printing devices are limited in the number of levels they can display. In contrast, a sampling device may generate a number of output levels (intensity values). As a result, multi-level halftoning is often used to convert a large number of output levels from a sampling device to a smaller number of levels that can be supported by a display device. For example, if the scanner could provide 256 levels of intensity values, but the printing device accurately displayed only 5 levels, then the multi-level halftoning system would replace a single 256 level value with It must be distributed to halftone cells containing a plurality of five-level pixels, so that when viewed with the human eye, the halftone cells appear as 256 level values.

In the multi-level halftoning method disclosed in the prior art, the input intensity value associated with the intensity value is determined by a threshold to determine the appropriate intensity level for each pixel in the multi-level halftone cell. It may be compared to the number of tables. For example, if each pixel is N
When represented as one of the levels, N-1 separate threshold tables may be used to generate the output image. In general, the comparison process is such that the comparison process is N-1
Is similar to that used for halftoning bitonality, except that it is repeated N-1 times for the threshold table of As in the case of monotonic halftoning, each threshold table contains a plurality of thresholds equal to the number of pixels in the halftone cell. As in the case of binary halftoning, the output of each comparison may have one digital bit, ie, a signal with either a logical "1" or a logical "0" value. Good. The output bit value indicates that the intensity value was greater than the threshold, ie, a logical “1”, or less than the threshold, ie, a logical “0”. In essence, the four intermediate matrices are four monotonic halftone cells. Elements of each intermediate matrix having the same coordinates are combined to form a 4-bit word. Each 4-bit word is then encoded to generate halftone output values for the pixels in the multi-level halftone cell. The resulting pixel values range from 0-4. That is, one level for each threshold matrix has one level for representing the presence of a pixel.

One application of the multi-level halftoning technique is the use of an alternative imaging technique whose spatial resolution is too low to achieve high image quality in binary halftoning due to physical and technical constraints. Improving usability and convenience. For example, it is known that full rendering of image details requires a screen with a frequency of at least 120 lines per inch. On the other hand, a smooth rendering that uses slow tone changes to reduce image quality requires at least 100 gray levels. If both conditions coincide, and if this is a traditional condition, then a binary device with at least 1200 dots per inch is required. Due to the technical limitations of rendering processes based on technologies such as thermal wax transfer, electrophotography and inkjet, binary devices with 1200 dpi resolution are not always possible, and multi-level halftoning requires
It provides a solution for improving the image quality in such devices. The use of the "multi-level" capability of these devices provides a logical solution to increase the number of reproducible tones and to improve image quality.

Multi-level halftoning techniques also provide a method for improving the image quality of rendering devices with relatively low spatial resolution. The use of multi-level halftoning as a way to overcome the image quality constraints of devices with relatively low spatial resolution implicitly implies that the spatial and tint resolution of the rendering device is, to some extent, "compatible". Based on assumptions.
Different classes of multi-level halftoning techniques are identified based on the number of microdots in a halftone cell driven in binary mode and at an intermediate level.

Multi-level halftoning also requires the cost of memory to render the image using the multi-level halftoning technique, using the same image with the same image quality, and higher resolution as the multi-level halftoning technique. Rather than rendering with a binary halftoning technique.

[0022]

The application of the unchanged binary or multi-level halftoning technique described above
It works well with printers that have properties that are close to theoretical ideals. In this case, the gray level increases linearly and the cell growth needs to be symmetric.
However, many typical output image reproduction devices, such as color laser printers, do not exhibit these characteristics. As such, using this technique without modification for such devices may result in some undesirable visual artifacts. For example, if the printer has a development threshold below which no toner is developed, these gray levels will not print and, as a result, will not produce separate gray levels for these cells. Another problem is that the gray levels of the cells do not increase monotonically, which can lead to unwanted contouring. Also, both of these problems have the potential to negate the benefits of using multi-bit halftoning schemes and suppress gray levels that are useful for printing.

Thus, there is a need for a system and method for improving conventional binary and multi-bit halftoning techniques by eliminating undesirable printing artifacts and maintaining the most gray scale levels. You.

[0024]

SUMMARY OF THE INVENTION A system and method for generating a halftone output image from a source image by an output image creation process is disclosed. The output image may have multiple halftone tables, where each halftone table has multiple halftone cells. Each halftone cell is associated with an array of pixel values determined using a threshold halftone table having at least one threshold halftone cell associated with the array of pixel locations. The source image may be represented as an array of intensity values. The system and method of the present invention involves defining a cell growth pattern for each threshold halftone cell in a threshold halftone table. The cell growth pattern includes an array of pixel locations within the threshold halftone cell. in addition,
The characteristics of the output image creation process affect the arrangement of thresholds for each threshold halftone cell in the threshold halftone table. The overall growth pattern for the threshold halftone table is also defined. The whole growth pattern
It has an array of pixel positions in the threshold halftone table. The characteristics of the output image creation process also affect the definition of the overall growth pattern. Each intensity value in the source image is associated with a halftone table in the output image. The pixel value of each halftone cell in each halftone table in the output image is set by comparing the associated intensity value to a threshold in the threshold halftone table.

The generation of a multi-bit halftone image on a display device from a continuous tone source image is improved by several methods described below. These methods of improving the quality of a multi-bit halftone image generated from a continuous tone source image may be employed separately or in concert.

For example, in the conventional method, the cell growth pattern is the same for all pixels in the halftone cell. For example, all halftone cells may exhibit a vertical growth pattern, where the first pixel may be grown in the upper left corner, for example, in each row of the halftone cell matrix. Down,
Further, the growth may be continued so as to shift rightward to fill a new row. Other types of vertical growth patterns are also possible, for example, may be grown upward and left. Alternatively, all halftone cells may exhibit a horizontal growth pattern. Here, the first pixel may be grown at the upper left corner of the halftone cell, for example, growing horizontally to the right in each column of the halftone cell matrix or down to fill a new column. You may continue. Finally, the halftone cell may represent a diagonal growth pattern. In addition, there are different types of horizontal and vertical growth patterns, just as there are different types of vertical growth patterns. In some conventional methods, each of these methods may be used individually to grow all halftone cells in a multi-bit halftone image.

In each of these various types of horizontal, vertical and diagonal growth patterns, the same number of pixels are darkened to the same level. However, due to the asymmetry of the electrophotographic process, each cell pattern will produce a slightly different gray value when measured by a densitometer. These cell patterns can be used in tables large enough that the repeating pattern can be maintained by tiling. In this way, the number of acquired gray levels grows, but the reference screen frequency is preserved. This repeating pattern is
In general, it is more effective at increasing the number of grayscale levels in printers such as electrophotographic or laser printers that have more image noise to provide randomization and consequently reduce the sharpness of repeating patterns. . As such, asymmetric cell growth techniques tend to be less effective in ink jets, which tend to reduce image noise (and often need to generate themselves, such as through air diffusion techniques).

In another refinement of the conventional method, different cells may be grown simultaneously. That is, one or more adjacent cells may start growing before one cell finishes growing. The growth of this discontinuous overlapping halftone cell differs from the growth of a conventional halftone cell. Conventional halftone cell growth is continuous, that is, once growth of one halftone cell begins, it does not stop until it finishes, and another halftone cell replaces the previous cell. Do not start growing until growth is over. As a result, at any given time, only one cell is in the process of growth. In the growth of discontinuous halftone cells, some halftone cells grow simultaneously and at different rates. Discontinuous growth results in different dot growth patterns, ie, different gray levels that result in the number of gray levels grown.

A third improvement of the conventional method of multi-bit halftone printing is developed by threshold effects. Here, some gray levels for pixel growth will not be printed because they are invisible. These threshold effects can occur as a characteristic of the threshold toner induced voltage of the imaging drum in an electrophotographic (eg, laser) printer. The traditional response to threshold effects has been to print the closest printable level with respect to the unprinted level. This correspondence has the potential to lead to unwanted contouring effects, for example when continuously changing gradients are printed. One method disclosed in the present invention for eliminating unwanted contouring due to threshold effects eliminates invisible gray levels and makes them the lowest (eg, the brightest).
It involves treating it as if it were equal to the gray level. Yet another method for removing unwanted contouring disclosed in the present invention addresses cases where pixel growth occurs very quickly. That is, the continuously raised gray level for pixel growth is output at a level that is much more saturated than usual. In this method for removing unwanted contouring in an oversaturated multi-level halftoning system, the same gray level is used before the next higher level can be printed.
Repeated over several levels.

These three methods may be used in combination in building a threshold array for a printer. These techniques may be applied by first characterizing the printer, as each type of printer is unique and cannot be easily conceived. The printer characterization process begins with the printing of a simple linear threshold array and a series of 256 grayscale patches. The gray scale values of these patches may be measured and displayed on the figure, and the dot growth pattern may be observed with a microscope. Analysis of the generated series of grayscale patches allows to discover typographical artifacts and the reasons for them,
It also improves the effect of printing artifacts and allows the selection of one or more of the above techniques to find the best threshold arrangement.

[0031]

Embodiments of the present invention will be described below with reference to the accompanying drawings. A method and system for improving a multi-bit halftone table in an image reproduction device such as a printer is disclosed. The following description is provided to enable any person skilled in the art to make and use the invention. For purposes of explanation, specific names have been provided to provide a thorough understanding of the present invention. Descriptions of specific applications are provided only as examples. Various modifications to the preferred embodiment will be readily apparent to those skilled in the art. Also, the general principles defined herein can be applied to other embodiments and applications without departing from the spirit and scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

One method known in the art for converting a continuous tone image to a halftone image is the halftoning process. This process is schematically illustrated in FIG. Input image 1410 typically consists of color or gray values from 0 to 255 levels at a particular resolution. The process of converting this to a halftone screen consists of the following steps. The input image is converted by the threshold table 1420 into a halftone image using methods known in the art as outlined in the prior art. This process is performed on a pixel-by-pixel basis until all pixels of the image have passed the threshold table, starting from the first pixel in the image and continuing to the next pixel. Each new transformed pixel is placed in a memory location called frame buffer 1430. Frame buffer 14
30 needs to be large enough to hold all the pixels of the page, or in the case of color, all the pixels in one color plane. After the page buffer is full, the image is ready for printing. The pixels are then sent to an imaging device 1440 of an output device such as a printer. In the case of a laser printer, the data is sent to a laser modulation circuit and processed in a raster fashion as known in the art.

Binary Halftone The binary halftone table defines the individual halftone cells that make up the halftone screen. Once the halftone cell is defined, each pixel location in the cell is given a threshold. One way to define the halftone cells is to create a table that defines the order in which the halftone cells will be filled. This is shown in FIG.
In this case, a 4 × 4 table 10 is defined. The table 10 has four halftone cells 20, each including a matrix of 2 × 2 pixel locations 30. Pixel position 3 in table 10
0 has an integer value between 1 and 16, respectively. Here, the position having a value of 1 corresponds to the filled table 10
, The position having a value of 2 is a filled second pixel position 30, and so on. As a result, the pixel location having a value of 16 is the last pixel location 30 that has been filled. Table 10 is defined as a cell growth table and defines a halftone screen.

In this example, the formed halftone screen has four spot functions. In this case, the spot function may be associated with the halftone cell 20. As for two of the spot functions, one is
Halftone cell 2 at upper left position of table 10
The other is associated with the halftone cell 20 at the lower right position of the table 10.
The second two spot functions, which are respectively associated with the upper right and lower left positions of the table 10, are filled after the pixel positions for the first two halftone cells are completely filled. For a halftone screen based on a pixel resolution of 600 dpi, the first two halftone cells grow at 45 ° with 212 spots per inch.

Once the cell growth table is defined, a threshold table 200 as shown in FIG. 2 can be created in a conventional binary halftoning system. In FIG.
The threshold has 16 integer values evenly distributed in the range between 1 and 256, although other values may be used as the threshold. One threshold 230 is associated with each of the 16 pixel locations. For example, the lowest threshold of 16 is located at the first pixel location 30 in the halftone screen 10 of FIG. 1 to be satisfied. The next lower threshold 32 is located at a second pixel location 30 in the halftone screen 10 of FIG. 1 to be met.
The threshold table continues to be developed in a similar manner until all pixel locations in cell growth table 200 are filled. As a result, FIG. 2 shows a complete threshold table. Here, each entry 230 in the table has a threshold value associated with switching pixel positions on during image rendering.

The manner in which this threshold table 200 is used to render an image is shown in FIGS. 3 (a) and 3 (b). FIG. 3 (a) shows a portion of the original image, where the image fragment is represented as a two-dimensional array 300A of intensity values. This image, shown in FIG. 3a, has light and dark horizontal stripes two pixels wide. The value 310A in the image display array 300A is known as the intensity value, or grayscale value. These intensity values are used with the threshold table 200 to reproduce the image using a binary halftoning technique.

The integer 50, which is the first value 310A in the image display array 300A, is compared with the integer 16, which is the first value in the threshold table 200. Since the gray level is larger than the threshold level, at the position 310B in the frame buffer 300B shown in FIG.
1 is arranged. Here, position 310B is associated with a value in image display array 300A. The second gray value 310A, integer 50, in the image display array is compared to the second threshold, integer 48, in table 200. In this case, the gray value 50 is also greater than the threshold 32, so that the frame buffer 30
In the next position in 0B, 1 is also arranged. The third through eighth gray values are below the corresponding threshold, so that the corresponding position in the frame buffer has 0
Is arranged. Thus, the frame buffer 30
0B is created as shown in FIG. This is a measure of the binary threshold halftone process as is well known in the art.

In the example shown in FIGS. 3A and 3B, the displayed image fragment is the size of one threshold table or one halftone screen. However, the image may be represented by a threshold table or an array of intensity values larger than a halftone screen. The threshold table may be used to generate halftone representations of these images by repeatedly applying the threshold table over an array of intensity values.

Multilevel Halftone The creation of a multilevel halftone table is an extension of just the binary case. In the case of multi-level halftone, the cell growth table as shown in FIG.
It is created and used in a manner similar to that described above for the binary halftone case. However, unlike binary halftoning, instead of a threshold table where each entry in the threshold table has a single value to determine whether a pixel is on or off, each entry in the table has multiple thresholds. A threshold table associated with is created. Each of the thresholds associated with the table entry (ie, pixel location) determines whether the gray value for the pixel location should be grown to the next level.

For example, in the specific case where the multilevel case corresponds to 16 levels (equal to 4 bits per pixel), 16 thresholds need to be created per pixel location. FIG. 4 shows an example of a table 400 for 16 levels per pixel threshold array. The first column 410 of the table is (at the top left of the halftone cell)
It has a first 16 levels from 0 to 15 corresponding to the first pixel position. The second column 420 corresponds to the second pixel location and includes the next 16 levels from 16 to 31, and so on.

A threshold table, such as the 16-level threshold table 400 shown in FIG. 4, may be combined with the cell growth table to form a halftone table. FIG. 5 shows an example of a 16-level halftone table 500 obtained by arranging the 16 columns of the threshold table 400 of FIG. 4 in a specific order.
Here, the specific order corresponds to the cell growth table of FIG.
It is defined by reading from left to right starting from the upper row and proceeding to the lower row. For example, as shown in FIG. 5, the first column of the table corresponds to pixel position 1, the second column of the table corresponds to pixel position 3, and the third column corresponds to pixel position 1. 9 and so on. The table reads each row horizontally and downwards in order, resulting in a table shown in FIG.
Is constructed from the cell growth table in.

FIGS. 6 (a) and 6 (b) show how
It shows whether the threshold table 500 of FIG. 5 is used with an image to fill the frame buffer. FIG.
(A) shows a part of the original image composed of bright and dark horizontal stripes. In FIG. 6A,
The image fragment is represented as a two-dimensional array 600A of intensity values. This image, shown in FIG. 6 (a), has light and dark horizontal stripes two pixels wide. Bright pixels in array 600A are associated with 40 intensity values, and dark pixels in array 600A are associated with 185 intensity values.

The process of filling the frame buffer for generating a display of an image using the multi-level halftone method is as follows. The first pixel value 40 at the position 610A in the image display array 600A is 1 in FIG.
The value is compared with the value 510 of the first sequence in the six-level threshold table 500. Since the forty gray levels or intensity values associated with the first pixel location are greater than all the levels in the corresponding entry 510 of the threshold table 500, a value of sixteen is used for the frame shown in FIG. The buffer 600B is located at a first position 610B. Next, the second pixel value 40 in the image is at position 620A and is compared with the second sequence value 520 in the threshold table 500 of FIG. In the second sequence value 520, the 40 intensity values correspond to the ninth level in the threshold sequence. As a result, the value 9
Is the second position 62 in the frame buffer 600B.
0B. The remaining entries in frame buffer 600B are completed in a manner similar to that described above, and frame buffer 600B is finished as shown in FIG. 6B. This describes the development of a multi-bit halftone table as is well known in the art.

Technique for Optimizing Multi-Bit Halftone Table FIG. 7 shows a first method for optimizing a multi-bit halftone table. In the conventional multi-bit halftone method known in the prior art, the permutation of the threshold is linear,
It is continuous. Due to the characteristics of the printing device, a conventional linear and continuous increase in the threshold may not be the best solution. The continuous permutation of the thresholds indicates that the printer allows printing at 16 gray levels per pixel. The fact may not be the case. For example, for a particular laser printer, there may be a threshold for printing, so that the first four gray levels of each pixel are not printed. If the threshold table contains these levels, the printer will not print these levels, and the output may produce unwanted artifacts, such as contouring. This can be corrected by constructing a threshold table that causes level skipping.

FIG. 7 is an example of a portion of a multi-level threshold table showing the use of a level skipping technique.
Taking the case where the first four gray levels at a pixel location are not printed, using level skipping, the non-print level is "skipped" by repeating the first threshold for each pixel location. obtain. As a result, the input image treats the first four levels of pixel positions as one gray level. That is,
Three levels will be skipped. This correction can be made at an early stage of cell growth or at 5% of the pixels in the cell.
Significant before 0% is formed. Since the use of level skipping only reduces the discrimination between intensity levels in the later stages of cell growth, level skipping correction is selectively performed for best effects to address the state of early cell growth. Can be adopted.

A similar characteristic appears in a printing device when more than 50% of the pixels in a cell are formed. When most pixels are formed,
Small changes in the threshold table do not have a noticeable effect on the output, and therefore these gray levels are wasted. The lack of contrast between the thresholds is assigned to the 800 portion of the threshold table shown in FIG. In this case, certain threshold levels are not used. For example, in the table fragment shown in FIG. 8, between levels 1 and 2 of pixel 16, 196, 197 and 19
The intensity value of 8 is skipped. Pixel 16 outputs at level 1 for 195 intensity values and at level 2 for 199 intensity values. This has the effect of repeating the same gray level several times. In this example,
One gray level is repeated for 196, 197 and 198 intensity values. This is called level repeating.

In all of the above cases, conventional pixel growth methods are acceptable. That is, the pixels in each halftone cell were continuously grown. As the first pixel begins to grow, it switches completely on, ie, grows until the intensity value of the image associated with the pixel exceeds the threshold associated with the highest gray level at the pixel location. The next pixel does not start growing until the pixel that started to grow before has completely turned on. In all cases, one pixel does not start until the previous pixel is completed. Due to the nature of the print engine, this conventional solution does not make it possible to obtain the maximum gray level.

An alternative to continuous pixel growth is non-continuous pixel growth. A discontinuous pixel growth technique is illustrated in part of the multi-level halftone table shown in FIG. In this case, the first five levels of the first pixel are grown for intensity values from 0 to 4. Then, the first five levels of the second pixel are:
While the growth of the first pixel is paused, it is grown for image intensity values from 5 to 9. After that, the next five levels (levels 6 to 10) of the first pixel are:
Grown for image intensity values from 10 to 14,
Further, thereafter, the image intensity value of 15-19 is paused. While the growth of the first pixel is temporarily stopped, the next level (6-10) of the second pixel location
Are grown for image intensity values of 15-19.

In the example of FIG. 9, the first five levels of the third pixel are then grown for image intensity values of 20-24. The process is, for a subset of levels,
The process is continued by alternately growing each of the three pixels or temporarily stopping the growth. The growth process ends with a first pixel, a second pixel, and
This is completed by terminating the growth of the third pixel.

There are several alternatives to this solution. One alternative is shown in FIG.
In this case, instead of completing the third pixel, the fourth pixel begins similarly, except that it begins to grow after the tenth level of the third pixel. The fifth pixel is
It begins to grow for five levels for image intensity values of 58-62, while the fourth pixel is in the process of growing and is temporarily stopped. In this case, as before,
There are pixels that start growing while other pixels have been completed. Other combinations are also possible and may be defined by the characteristics of the printer and the characteristics of the source image.

All of the above-described methods are useful in generating a halftone table in that they increase the number of gray levels generated using the smallest cell size.
It is an improved method with good results. This is because the smallest cell size results in the highest screen frequency,
Optimal. Higher screen frequencies reproduce more detail in the image and hide the fact that the image is made from a halftone screen. A disadvantage to the use of small cell sizes in the prior art is that it reduces the number of gray levels. The main advantage of multi-level halftoning is that without increasing cell size,
That is to increase the number of gray levels. The above method further improves the multi-level halftoning technique by addressing the halftoning technique to further increase the number of gray levels.

FIGS. 11 (a)-(d) and FIG. 12 also illustrate another method of increasing the number of gray levels that works for binary with a multi-level halftone table. FIG. 1 shows an example of the dot growth pattern. In this pattern, the second pixel is
It was placed adjacent to and horizontal to the first pixel. This is an example of growth in the horizontal direction. FIG. 11B
Shows different examples of dot growth patterns. FIG.
In 1 (b), the second pixel is arranged at a position perpendicular to the first pixel. This can be characterized as vertical growth. The dot growth patterns in (c) and (d) of FIG. 11 are two examples of oblique growth. All four of these cell growth patterns can be incorporated into a larger cell pattern as shown in FIG. Each of these patterns has the same number of filled cells, but due to the asymmetric nature of the electrophotographic process, each cell pattern produces a slightly different gray value when measured by a densitometer. Will. If these cell patterns are used in a table large enough that the repeating pattern can be maintained by tiling, a reference screen frequency is reserved, except that an increase in the number of gray levels is obtained Will.

These methods can be used together in constructing a threshold array for a printer. Because each type of printer is unique and not easily predictable, it has been found that the best way to determine how to apply these techniques is to first characterize the printer. . This consists of printing a series of 256 grayscale patches, starting with a single linear threshold array. The gray values of these patches are measured and plotted. Thereafter, the shape of the curve is analyzed and the dot growth pattern is observed with a microscope. This allows you to find artifacts on the print and understand why they occurred. This is needed to develop the means by which the technique is used to remove artifacts and to develop the best threshold arrays. FIG. 13 shows the results of plotting the measured optical densities of the 256 gray patches using a linear threshold table. From this, it can be observed that many gray patches have the same optical density. This is the level skipping,
It indicates that these gray levels are appropriate so that they are not repeated. Also, at some gray levels, it can be observed that there is a jump in optical density. This means that more gray levels are needed at this density. This can be achieved by a simultaneous dot growth method or by asymmetric dot growth. Both methods have been tried and the best result can be selected to use.

The present invention is not limited to the illustrated embodiment, and it goes without saying that various improvements and design changes can be made without departing from the spirit of the present invention.

[0055]

As is evident from the above description, the present invention eliminates undesired printing artifacts and maintains the most gray scale levels, thereby reducing the conventional binary and multi-bit half-tones. It is possible to improve the toning technique.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a cell growth table used in a binary halftoning system.

FIG. 2 is a view showing an example of a threshold value table used together with the cell growth table shown in FIG. 1;

FIG. 3A shows an example of a continuous tone image represented as an array of intensity values. FIG. 3B is a diagram showing an example of a frame buffer generated by converting the array of intensity values shown in FIG. 3A using the threshold table shown in FIG.

FIG. 4 is a diagram showing an example of a multi-level threshold table indicating 16 gray levels per each pixel position.

FIG. 5 is a diagram showing an example of a multi-level threshold table used together with the cell growth table shown in FIG. 1; Here, each pixel position indicates 16 gray levels per pixel position.

FIG. 6A shows an example of a continuous tone image represented as an array of intensity values. FIG. 6B is a diagram illustrating an example of a frame buffer generated by converting the array of intensity values illustrated in FIG. 6A using the threshold value table illustrated in FIG. 5;

FIG. 7 is a diagram showing an example of a part of a multi-level threshold table showing a level skipping technique used in an embodiment of the present invention;

FIG. 8 is a diagram showing an example of a part of a multi-level threshold table showing a level repetition technique used in an embodiment of the present invention.

FIG. 9 is a diagram showing a first example of a part of a multi-level threshold table showing a discontinuous pixel growth technique used in an embodiment of the present invention;

FIG. 10 is a diagram showing a second example of a part of a multi-level threshold table showing a discontinuous pixel growth technique used in an embodiment of the present invention.

FIG. 11A is a diagram showing a first example of a cell growth table for 2 × 2 halftone cells used in an embodiment of the present invention. (B) A diagram showing a second example of a cell growth table for 2 × 2 halftone cells used in the embodiment of the present invention. (C) A diagram showing a third example of a cell growth table for 2 × 2 halftone cells used in the embodiment of the present invention. (D) A diagram showing a fourth example of a cell growth table for 2 × 2 halftone cells used in the embodiment of the present invention.

FIG. 12 is a diagram showing an example of an 8 × 8 cell growth table which combines various cell growth patterns shown in FIGS. 11 (a) to 11 (d).

FIG. 13 is a graph showing measured optical densities of a series of 256 gray scale patches.

FIG. 14 is a block diagram showing a halftoning system used in one embodiment of the present invention.

[Explanation of symbols]

10 Table 20 Pixel Position 30 Halftone Cell 200, 400, 500 Threshold Table 230 Threshold 300A, 600A Image Display Array 310A Gray Value in Image Display Array 300B, 600B Frame Buffer 310B, 610B, 620B Position in frame buffer 410 First column of table 420 Second column of table 610A, 620A Position in image display array

Continued on the front page F term (reference) 2C262 AA02 AA04 AA24 AA26 AB13 BB03 BB22 BB25 BB27 BC01 5B057 CA08 CA16 CB08 CB16 CC02 CE11 CH07 5C077 LL19 MP01 NN02 NN08 PQ23 RR09 RR16

Claims (20)

[Claims] 1. A method for generating a halftone output image from a source image by an output image creation process, the output image having a plurality of halftone tables,
Each halftone table has a plurality of halftone cells, and each halftone cell is determined using a threshold halftone table having at least one threshold halftone cell associated with an array of pixel locations. Defining a cell growth pattern for each threshold halftone cell in the threshold halftone table, wherein the source image is represented as an array of intensity values. The cell growth pattern having an array of pixel locations within the threshold halftone cell, and wherein characteristics of the output image creation process affect the array of thresholds for each threshold halftone cell in the threshold halftone table; All growth patterns for the above threshold halftone table A step of defining, the total growth pattern has a sequence of pixel positions within the threshold halftone table,
Also, the step that the characteristics of the output image creation process affects the definition of the entire growth pattern; the step of relating each intensity value in the source image to the halftone table in the output image; Setting a pixel value of each halftone cell in each halftone table in the output image by comparing with a threshold value of the threshold halftone table. Generation method. 2. The method for generating a halftone output image according to claim 1, wherein a binary halftone output image is generated. 3. Each pixel value has a single bit,
3. The method according to claim 2, wherein each pixel position has a binary threshold. 4. The method for generating a halftone output image according to claim 1, wherein a multilevel halftone output image is generated. 5. The method of generating a halftone output image according to claim 4, wherein each pixel value has a multi-bit, and each pixel position has a plurality of multi-level thresholds. 6. The method according to claim 5, wherein the plurality of multi-level thresholds at the pixel positions have continuous intensity values. 7. The method according to claim 1, wherein the output image is generated on a digital imaging device. 8. The plurality of threshold halftone cells having the threshold halftone table employ a plurality of unique arrangements, wherein each arrangement is such that the pixel position in each threshold halftone cell is in a unique direction. 2. The method according to claim 1, wherein the halftone output image is set. 9. A method for generating a multi-level halftone output image from a source image by an output image creation process, wherein the output image has a plurality of halftone cells, each halftone cell being a multi-bit halftone cell. Defining a threshold array having a plurality of halftone cells, the source image being represented as an array of intensity values, wherein each multi-bit pixel value is associated with an array of pixel values. Is associated with a plurality of thresholds, and the characteristics of the output image creation process affect the threshold definition associated with each multi-bit pixel value associated with a halftone cell in the threshold array, and The threshold value of adjacent multi-bit pixels in the threshold array is different from that of adjacent multi-bit pixels in the threshold array. Associating each halftone cell in the output image with a halftone cell in the threshold array; and converting each intensity value in the source image to a plurality of multiple bit pixels in the output image. Setting the pixel value of each halftone cell in the output image by comparing the associated intensity value with a threshold in a threshold array. To generate a multi-level halftone output image. 10. The step of defining the threshold array further comprises omitting one or more of a plurality of thresholds associated with the pixel values of the multiple bits in the threshold array, Omission,
The method of claim 9, further comprising the step of being affected by characteristics of the output image creation process. 11. The step of defining an array of thresholds further comprises the step of repeating thresholds at a plurality of thresholds associated with the multi-bit pixel values at the thresholds, wherein the repetition of the thresholds is characteristic of an output image creation process 10. A method according to claim 9, comprising the step of being influenced by: 12. A method for generating a multi-level halftone output image from a source image by an output image creation process, wherein the output image has a plurality of halftone cells, each halftone cell comprising a multi-bit halftone cell. Defining a threshold array having a plurality of halftone cells, the source image being represented as an array of intensity values, wherein each multi-bit pixel value is associated with an array of pixel values. Is associated with a set of thresholds, and characteristics of the output image creation process affect the definition of a set of thresholds associated with each multi-bit pixel value associated with a halftone cell in the threshold array. And associating each halftone cell in the output image with a halftone cell in the threshold array. Associating each intensity value in the source image with a plurality of multi-bit pixel values in the output image; and comparing the associated intensity value with a threshold value in the threshold array to obtain each halftone in the output image. Setting a pixel value of a cell. A method for generating a multi-level halftone output image, the method comprising: 13. The step of defining the threshold array having a plurality of halftone cells, wherein each multi-bit pixel value is associated with a series of thresholds,
13. The multi-level halftone of claim 12, further comprising the step of repeating the sequential threshold, wherein the repeating of the sequential threshold is affected by characteristics of an output imaging process. How to generate the output image. 14. The method of claim 11, wherein each multi-bit pixel value is associated with a series of thresholds, wherein defining the threshold array with a plurality of halftone cells further comprises:
The method of claim 1, further comprising the step of omitting one or more sequential thresholds, wherein omitting the sequential thresholds is affected by characteristics of the output image creation process. 13. The method for generating a multilevel halftone output image according to item 12. 15. An output image data structure having a plurality of halftone cells each associated with an array of multi-bit pixel values in a system for generating a halftone output image from a source image by an output image creation process. A threshold array having a plurality of halftone cells associated with an array of pixel locations, wherein each pixel location is associated with a series of thresholds, the thresholds of the threshold array having a continuous value; Also, the characteristics of the output image creation process affect the definition of the set of thresholds associated with each pixel location, and furthermore, the threshold that each halftone cell in the output image is associated with a halftone cell in the threshold array. A source image data structure having an array and an array of intensity values, each intensity value being included in an output image data structure. The multi-bit pixel value in the array associated with the halftone cell in the output image data structure and associated with the halftone cell in the output image data structure determines the associated intensity value as
A source image data structure defined by comparison with a series of thresholds at pixel locations of associated halftone cells of the threshold array. 16. The system of claim 15, wherein the sequential thresholds are repeated and the sequential thresholds are affected by characteristics of the output image creation process. . 17. The system of claim 15, wherein the sequential thresholds are omitted and the omission of the sequential thresholds is affected by characteristics of the output image creation process. 18. The system according to claim 15, wherein the series of thresholds associated with the pixel locations in the threshold array is non-continuous. 19. A system for generating a halftone output image from a source image by an output image creation process, the output image data structure having a plurality of halftone tables, each halftone table comprising a plurality of halftone tables. A threshold halftone table comprising an output image data structure associated with an array of pixel values and at least one threshold halftone cell associated with an array of threshold values, the halftone table including a tone cell; Thus, each threshold halftone cell in the threshold halftone table is associated with a halftone cell in each halftone table in the output image data structure, and each pixel value in each halftone cell in the output image data structure is In the associated threshold halftone cell A threshold halftone table associated with a threshold, and a cell growth pattern for each threshold halftone cell in the threshold halftone table, the cell growth pattern having an array of thresholds in the threshold halftone cell; A cell growth pattern in which the characteristics of the output image creation process affect the arrangement of thresholds in the threshold halftone cell; and a full growth pattern for the threshold halftone table, wherein the total growth pattern is a threshold halftone table. A source image data structure having an array of intensity values and an entire growth pattern wherein the characteristics of the output image creation process affect the array of threshold values in the threshold halftone table; Each intensity value is a half-toe in the output image data structure. A source image data structure associated with the table and wherein the pixel values in each halftone table in the output image data structure are defined by comparing the associated intensity value to the associated threshold. And a system for generating a halftone output image. 20. A system for generating a halftone output image from a source image by an output image creation process, the output image data structure having a plurality of halftone tables, each halftone table comprising a plurality of halftone tables. An output image data structure including a tone cell and each halftone cell associated with an array of pixel values; and at least one threshold halftone cell associated with an array of two or more thresholds. A threshold halftone table, wherein each threshold halftone cell in the threshold halftone table is associated with a halftone cell in each halftone table in the output image data structure, and each halftone cell in the output image data structure. Each pixel value in the tone cell is associated with an associated threshold A threshold halftone table that is associated with a series of thresholds in the halftone cell; and a cell growth pattern for each threshold halftone cell in the threshold halftone table, wherein the cell growth pattern defines an array of thresholds in the threshold halftone cell. Having a cell growth pattern in which the characteristics of the output image creation process affect the arrangement of the thresholds in the threshold halftone cell; and a total growth pattern for the threshold halftone table, wherein the total growth pattern is A source image data structure having an array of pixel values in a threshold halftone table and having a full growth pattern in which characteristics of the output image creation process affect the array of thresholds in the threshold halftone table; and an array of intensity values Where each intensity value is the output image The pixel values in each halftone table in the output image data structure are defined by comparing the associated intensity values to an associated set of thresholds. And a source image data structure.