JPH07306940A - Method for conversion of color bit map into monochrome data - Google Patents

Abstract

PURPOSE: To provide a processor unit for converting source image data into a rastered monochrome bit map. CONSTITUTION: A processor unit 10 is provided with a random access memory 14, and a storage memory 16 including a gray scale conversion table 28, dither matrix 30, procedure for judging the target image of a dither tiling position, scaling conversion procedure 36, and sub-routine for establishing a gray scale value array 24 indexed to a bit mask array 26. The processor unit is provided with a conversion cycle for executing a method for converting source image data into a rastered and compressed monochrome bit map, and converting a source pixel into a gray scale value and output position bits in the two arrays. When all the source pixels are converted into the same gray scale value, a pre-reading cycle is started, and while the data byte of the unconverted source image is as long as that of the source image converted in the conversion cycle, source pixel lines are pre-read. A dither processing is operated to the converted output position bit of the converted source pixel.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a process for converting a source of bitmap data into monochrome data, and more particularly, to process bitmap data according to resolution without repeatedly processing even one source data. Related to converting to scaled monochrome data.

[0002]

BACKGROUND OF THE INVENTION The prior art process of converting a source image, usually color source data, into a rasterized monochrome bitmap typically examines each pixel of the source bitmap. , It is required to convert the color information for each pixel into grayscale values. After the conversion, the correct dither matrix is retrieved to produce scaled output pixels.
This process is slow, and involves a lot of calculation, which is troublesome. This feature of the process allows visual color image data, such as those used in laser printers with print engines capable of 300 dots per inch or even 600 dots per inch, to be converted into high resolution raster bitmaps. This is especially noticeable when converting to. Rasterized bitmaps for monochrome images at a resolution of 300 dots / inch are approximately 1 per letter size page.
Requires megabytes of raster memory. 600 dots /
Inches, 3.8 megabytes of raster memory is required. The rasterized bits of the output image, where this transformation process transforms each unique source pixel one at a time and performs dithering on each scaled group of output pixels resulting from each source pixel. To
To process 3.8 megabytes, 300 million conversion calculations and 3
Billion dithering steps are required and can be quite time consuming.

Laser printers typically include on-board or internal memory that can receive rasterized data bitmap information from a host processor. This bitmap is used to create an intermediate strip page of rasterized data.
The intermediate strip page of rasterized data is sent to the laser print engine, which operates at a constant speed. If new rasterized data is not available at a rate that keeps up with its engine operation, an "overrun" print occurs and the page cannot print.
To avoid this happening, intermediate page formatting must occur fast enough to keep up with the print engine. This means that if the host processor is sending data at a fast enough rate, or if the printer is sending it from the host processor to internal memory, perhaps slower, then the destination bitmap
Achieved only with enough internal memory to hold a sufficient portion of the image. Conversion, dithering, and rasterization execution time in the printer's random access memory (RAM) or host processor is longer than the time the intermediate strip page sent to the print engine should be created. , Print overrun occurs.

[0004] Prior art conversion procedures often have long run times and have been unable to keep up with the build time required to create intermediate strip pages for transmission to the print engine. as a result,
Many data conversion and compression techniques are commonly used in host processors, prior to the data transfer to the printer memory to increase the data transfer rate from the host processor to the memory on the board, Intermediate page creation is initiated during the transfer, while at the same time minimizing the internal printer RAM requirements for raster memory.

Therefore, in the prior art, the conversion of a color image source into a rasterized monochrome bitmap image is usually completed and stored in the host processor RAM. And the compression procedure is to transfer the bitmap image data on the printer board.
Executed before transferring to RAM.

What is needed is this conversion by eliminating the need to convert the color data information of each unique pixel into a grayscale value when a string of repeating color data is present. It is a process that speeds up the process. Also, what is needed to improve run-time speed is derived from the transformation of source pixels with repeating parts, especially when upscaling to a higher resolution image than the original source image. , Which eliminates the need to dither the scaled output pixels. Achieving either or both of these objectives facilitates the conversion process and the use of device-independent monochrome bitmap creation capabilities in either the host processor, printer processor, or peripheral device processor. To

[0007]

SUMMARY OF THE INVENTION It is therefore an object of the present invention to operate on a string of source pixels with repeating parts, which need not all be converted to grayscale values, and All of the scaled output pixels derived from scaling are to provide processing that does not necessarily have to be individually dithered to produce an all-rasterized bitmap.

[0008]

A processor unit is used to convert source image data into a rasterized monochrome bitmap, a random access memory (RAM), a processor, and a source. Includes a control memory that holds the multiple steps required to convert the image data to a rasterized monochrome bitmap. This control memory is
Grayscale conversion table, dither matrix, procedure for determining target image dither tiling position, scaling conversion calculation, and sub-routines for establishing two arrays, grayscale value array and bit mask array , Each of which has an output image dither
It has many elements equal to the dither pattern width of the tile.

The processor unit is used in connection with three conversion processes, including an initial conversion cycle, a look-ahead cycle, and an optional compressed data dither cycle.

The process begins with an initial conversion cycle, where data information, usually color data, for the source pixels is converted to grayscale values that are added to the elements of the grayscale value array. The appropriately scaled output position bits are then added to the corresponding bit mask elements. The conversion of source pixels and the addition of output position bits to the corresponding bit mask element, or additional bit mask array element, causes the counter to fill the dither pattern tile width or row in the output image. Will continue until it is determined that there are sufficient output position bits. The first element of the bit mask array is then examined to determine if it contains all of the output position bits, and if so, the element is normally a color with repeated parts.・ Data,
Indicates that repeated data exists.

If the above inspection shows repeated data in the source image, a look-ahead cycle is initiated in which the unconverted data information for the additional source pixels is read and directly the unconverted for the already converted source pixels. , The read ahead cycle is continued until either the end of the source line is reached or a different value of source data appears. The look-ahead cycle ends and the dither cycle begins when the look-ahead reaches the end of the row or a source pixel with a different data value appears.

The inspection of the first bit mask array element is
The read-ahead cycle is skipped, and processing proceeds directly to the dither cycle, indicating that not all output position bits are located in the first bit mask element.

The dither cycle begins with the selection of the appropriate dither matrix for the first element of the grayscale value array.
The selected dither matrix is the actual position X and position in the target image dither tile for the first output position bit to be dithered, with the first bit in the dither matrix at position X and position Y. It is rotated or repositioned to correspond to Y. Then, the entire first element of the bit mask array is ANDed with the rotated dither matrix,
Generate output bits. Each bit mask element corresponds in size to the output image dither tile width,
For data compression purposes, it is divisible by 8 on the X-axis to facilitate direct data compression.

If repeated data occurs, dithering computes the number of scaled output bits in the repeated string and produces output data, or first.
Try to compress the dither output signal obtained from the AND operation of the bit mask elements of and the dither matrix into a single byte,
If successful, it produces a repeated run compressed data signal for transfer to the output buffer. If the repeated run compression string cannot be generated, a character run compression data signal is generated and transferred to the output buffer.

If read ahead does not occur, essentially the same process is used to AND the bit mask array elements that follow the appropriate rotation dither matrix and OR the outputs together to form the output signal. To do.

[0016]

DETAILED DESCRIPTION OF THE INVENTION In the following, the invention will be described in connection with the conversion of color video image data to monochrome rasterized image data for a laser printer. However, it is understood that any source data and other devices such as plotters, facsimile machines, etc. can use the procedures implemented within the scope of the present invention.

In FIG. 1, the processor unit 10 includes a microprocessor 12, an input port 18, a random access memory (RAM) 14, a storage memory 16, and an output port 20.
, All connected by bus 22 and including all of the steps necessary to ensure that the color image data bitmap is converted to a scaled monochrome bitmap. The processor unit 10 is not limited to a host processor,
It may be incorporated into a printer or a variety of different data processing units, including independent stand-alone peripheral units.

Image data from the input device is received via input port 18 and read into RAM 14. This image data can take one of several forms. This image data is a sub-image,
Or a pixel, where the descriptor is adjacent to each sub-image and is processed with each sub-image, or the descriptor of the sub-image or pixel is (connected to) the actual sub-image data. It is received separately. In both cases, when the processor unit 10 receives the image data, the image data is parsed, the commands and data are processed, the source image color data and the source image data are processed.
Raster graphics data is derived that includes a position descriptor for each pixel of the image.

The storage memory 16 is the processor unit 10.
Determines the grayscale conversion table 28, the dither matrix 30, the target dither tiling position, even when is used in an integrated configuration with the host processor, or even when the processor unit 10 is a standalone peripheral unit. Procedure 34, a scaling conversion procedure 36, two arrays of a grayscale value array 24 and a bit mask array 26, and a subroutine for performing a conversion procedure including a conversion procedure 32.

The two arrays, grayscale value array 24 and bit mask array 26, each contain a number of elements equal to the dither pattern width of the output image dither tile. The two arrays are each divisible by 8 to facilitate data compression in the preferred embodiment.

In the preferred embodiment, the grayscale monochrome data is contained in a single 8-bit byte, thus between the white represented by the binary number 255 and the total black represented by the binary number 0. It provides 254 gray scale display functions. In all cases, regardless of whether the color data is contained in a single byte format, or the color data format is 4 bits / pixel, 24 bits / pixel, or some other number of bits, for example: The colors that can be displayed in 16 colors in the 4-bit / pixel color data format are spread over the spectrum of monochrome data that can display 256 gradations, so this source color data is an appropriate monochrome gray. It must be normalized while being converted to grayscale.

The bit mask array 26 holds information indicating the relative position of the output pixels generated with respect to each associated gray scale. Similar to the grayscale value array,
The bit mask array contains a number of elements equal to the width of the dither pattern matrix, which in the preferred embodiment is divisible by 8 to facilitate data compression. Each element of the bit mask array contains a block of bits of sufficient size to hold the number of output position bits equal to the width of the dither pattern tile of the output image, so that all of the output pixels are Same grayscale value, they are contained in a single bit mask element. The elements of the grayscale value array 24 and the bit mask array 26 are associated with each other by their element indices.

To illustrate and describe the preferred embodiment,
The following parameters are used. That is, the processor unit 10 is mounted in a laser printer, and the source color image is a video image of the host processor with rasterized pixels of 74 / inch resolution, and the source color of each pixel is・ Data is white,
And the dither pattern matrix for the output image tiles contained in a single 8-bit byte corresponding to 256-color displayable colors including black and black (X _DN and Y _DN , where The DN is in the range of DN = 1 to DN = 16), the device to which the output pixel data is sent is the print engine of a laser printer with a raster engine with a resolution of 600 dots per inch (DPI), The source image typically contains a significant amount of repetitive data.

Referring now to the flow diagram of FIG. 2, the actions taken at the peripheral unit to convert the pixels of the source color image into a monochrome bitmap are shown. In the first step, the source image data is received via the input port 18, converted to a raster graphic bitmap, the source color data format is identified and the scaling format used is scaled. Calculated using the conversion process. With respect to the scaled format, the example source format is 74 pixels / inch and the output format, or resolution, is 600 DPI, which allows for each input pixel horizontal or X direction and vertical or Y direction. Produces a scaling factor of approximately 8.1 output pixels in both directions of, thus, for each 74 input pixels
Periodic rounded scaling is required to provide an output of 600 output pixels. In FIG. 2, the number of X pixels produced by scaling is shown as X _S , the number of Y rows produced by scaling is shown as Y _S , and in both of these embodiments, the periodic rounding is S depending on whether it occurs
Is in the range from S = 1 to S = 8 or 9.

The number of destination pixels, or output pixels, produced by each source pixel is proportional to the number of source pixels in the corresponding source row divided by the number of output pixels in a raster row, so that the extra Useful for precise conversion where output pixels are added to the image at unique positions in each row. For this reason, all scaled calculations,
And the addition of extra output pixels starts anew at the beginning of each row and is measured for each row from a fixed reference point.
Similarly, all scaling calculations and additions of additional output rows are started from fixed reference points. This feature allows the input and processing of source color images in a fragmented or segmented manner.

In addition to scaling for different resolutions, it should be pointed out that scaling that can occur for image sizing, ie both upsizing and downsizing, is also possible. In any case, the principle implemented in this method is the same.

As shown in FIG. 2, the source image data is received from block 18 and then block 44.
At, the total number of input image lines and the length are determined. Next, at block 46, the total number of output lines and the length are determined. The scaling factors X _S and Y _{S for the} X and Y axes are then determined at block 48. At block 50, the starting output image line and the number of lines to be processed for a fixed reference point for the output image are determined. The first byte of the source color data information is then converted to the monochrome grayscale value determined from the grayscale conversion table 28, as shown in box 52 of FIGS. This grayscale value is then added to the first element of the array of grayscale values as illustrated in element 200 of FIG. 7 and box 54 of FIG. Then, in this example, the first source
The scaled output position bits, which are the eight output position bits for the pixel, are stored in the 16 bits of the bit mask element 201 of FIG. 7, one at a time, as shown in box 56 of FIG. Is added to the first bit mask array element 201 at the appropriate position in. As shown in FIG. 2, after each output position bit is added to the bit mask element, each output position bit is determined at decision box 58 to determine if the end of the raster line of the output image has been reached. To be inspected. Furthermore, each output position bit is added to the bit mask element so that the first bit mask array is sufficient to exactly fill the dither pattern tile width of the output image (16 bits wide in this example). Each output position bit is counted in decision box 60 to determine when to include the output position bit.

At this time, the decision box 60 counts only the 8 output position bits in the first bit mask element, so that the next source
The color data bytes for the pixel are retrieved,
Converted to its grayscale value. Next, in decision box 64, the grayscale value is checked to determine if it is equal to the grayscale value of the first source pixel in box 54. Upon inspection, the grayscale value is boxed with respect to the first source pixel.
If determined at 52 and shown to be different from the value stored at box 54, then the second grayscale value is the second element of the array of grayscale values, as shown in FIG. 7 and box 68 in FIG. Added to 202, the appropriate number of scaled output position bits is added to the corresponding second element 203 of the bit mask array, as shown in FIG. 7 and box 70 of FIG.

The number of output position bits representing the resulting output pixels is increased until an appropriate number is added equal to the width of the dither pattern tile for the monochrome output image.
It is continuously counted in decision box 60 as being added to any element of the bit mask array. If the scale factor is low (eg, 4 output position bits per source pixel), then the conversion of the source pixel will cause each new gray occurrence to occur until a full dither pattern tile width of output position bits has been generated. Continue adding the scale value to the new grayscale value array element and adding the scaled output position bits to the corresponding bit mask array element. In addition, each source
As the color data for a pixel is converted to a grayscale value, that value is examined in decision box 64 to determine if that grayscale value has occurred before the current conversion cycle. If so, this grayscale value will not be added back to the next element of the grayscale value array, and instead the output position bit associated with the converted source pixel will be
As shown in box 66, it is added at the appropriate relative position within the bit mask element that is already associated with the previously encountered grayscale value.

At this time, the process of converting the source color data into gray scale values is stopped. The process stops even if the scaling process requires additional output position bits for the particular source pixel that was converted. If the result of the scaling is an unused output position bit for the last converted source pixel and no look ahead processing has occurred, this source pixel will be The remaining scaled output position bits are reused as the transformed first new source pixel.

If no scaling process has occurred, that is, the ratio of input pixels to output pixels is 1: 1 and each source pixel has a different color and is converted to a different grayscale, 16 Elements are required in both the grayscale value array 24 and the bit mask array 26. Thus, each array has a sufficient number of elements to fill the output dither pattern tile width. However, it is believed that not all elements are used in any transformation and filling of the array.

Similarly, it is more likely that only one or two grayscale values will occur in any conversion cycle, so that each bit mask element is also fully dithered.
It has a bit block large enough to hold the pattern tile width.

As shown in FIG. 7, when the bit mask element of the array contains the correct number of output position bits, the first bit mask element 201 will be replaced by its first bit mask element 201, as shown in decision box 72 of FIG. Checked to see if all 16 bits are on. Not all of them were on in the first example of the conversion cycle, but if they were all on, it means that the two first source pixels examined in this example are the same color. Will be shown. Since not all bits of element 201 of FIG. 7 were on, the conversion process proceeds to dither the output position bits for the final transfer to the output memory buffer, as shown in box 96.

As shown in box 74 of FIG. 4, this dithering process involves the first element 20 from the grayscale value array.
Start by choosing the correct dither matrix for 0. Once the correct dither pattern matrix is selected, the position Y _n in the output image dither pattern tile for the particular output position bit to be dithered is determined and the dither The same position Y _DN in the pattern matrix is selected. Similarly, as shown in box 78, the bit offset in the destination bitmap dither tile for position X _N , the first output position bit to be dithered, is calculated and the dither pattern matrix selected. The row of Y _DNs generated is rotated in box 80 to make the position X _DN of the first output position bit from the dither matrix correspond to the position X _N of the first output bit in the destination dither tile.

Thereafter, as shown in box 82, the bit mask element corresponding to the particular grayscale value element (bit mask element 201 in this example) produces what will ultimately be the output pixel. Is then ANDed with the rows of the rotated dither pattern matrix. The bit mask element 201 is then OR'ed with a test flag to determine if all 16 output position bits are on, as shown in decision box 84. If not all are on, the process is repeated for the next grayscale value and the next grayscale array element 202 for the associated output position bit.
This process continues with the grayscale value array and the next element of the bit mask array until all 16 bits in the test flag are turned on, thus producing 16 bits of output data, Compression is attempted as shown in FIG. 6 to calculate the number of output bytes ultimately sent to the output buffer in box 96, 8 in box 86.
Divided by. No provision is made to store any remaining output bits, or pixels, but all of the output pixels resulting only from the bit mask elements of the converted source color data are as described above. Divided by 8, there is nothing for this group of output pixels.

If data compression is not performed as in FIG. 5, the output pixels are sent directly to the output buffer where the output pixels are stored and used to create a complete raster row as shown in box 96. . Once a complete raster line has been created, it will be output to the printer via output port 98.
Sent to the engine.

If data compression is performed, the next step attempts to compress the output data as shown in FIG. This groups the 16 output pixels into output bytes, as shown in box 88, and compares those bytes to determine if they have the same value as shown in decision box 90. It is achieved by comparing with each other. In this particular example, the output bytes have different values because they are derived from different dither pattern matrices. As a result, in box 94, a character run compressed data string is created,
As shown in box 96, it is sent to the output buffer used in creating the full raster line command. Once the full raster line command is created, it is sent to the printer via output port 98. Since there is no output pixel remainder, the remainder from box 122 is not sent to the first bit mask element 201, as shown in box 56 of FIG. 26 All elements in both are reset to 0.

In practice, the use of the above conversion process with a grayscale value array 24 and a bit mask array 26 to scale the output position bits that do not have any repeating parts in the data is typically found in the prior art. That's a 30% time savings.

It is also possible to save more time by using the present invention when the source data has repeating parts. In these situations, what is needed to convert all the source color data for repeating pixels, i.e. each source with repeating parts.
Anything needed to dither the output pixel associated with the pixel is eliminated.

In the first conversion cycle, the color data for the two source pixels is converted, resulting in the two elements of the bit mask array used to create the overall 16-bit output position map. become. In the next conversion cycle of this example, the next source pixel, the third pixel, is the raster image of the source image.
It begins with a string of repeating pixels of the same color that extend across the line. Grayscale source data
The cycle of converting to data is converted to a grayscale value in box 52 of FIG. 2 and added to the element 200 of the grayscale value array 24 in box 54 as shown in FIG. • At the next source pixel to become a pixel (ie the third pixel), it is newly started as shown in FIG. After adding the grayscale value to element 200, the scaled output position bits are added to bit mask element 201. This process is repeated for the next pixel, in this example the fourth source pixel is converted to a grayscale value in box 62 and added in decision box 64 to FIG. 8 and element 200 in box 54 of FIG. Are checked against the grayscale value. If they have the same value, i.e. they have the same color data, then the appropriate number of scaled output position bits for the fourth source pixel is also added to the bit mask element 201, thus Bit mask element 2
Fill 01.

The count in decision box 60 that determines if there is a full dither pattern width of output position bits is such that all bits in the first bit mask element are on, as shown in decision box 72. The test is started to determine if it is. If all bits in the first bit mask element are on, all final output bits will be the same grayscale.

At this point in the process, instead of proceeding to dithering the scaled output position bits, the remaining unconverted source color data is sourced as described in the first conversion cycle of this example. A check procedure is started to determine if additional pixels in a row can be used to see if they are the same color, and thus if the conversion process continues, it will be converted to the same grayscale value.
By using the original source color data,
There is no need to convert each source pixel into a grayscale value, thus considerably reducing the calculation time.
Furthermore, as will be described later, the scaling operation and the dither processing are integrated to further save time.

The checking procedure begins in decision box 100 with the determination of whether a single byte of source color data contains color information for more than one source pixel. If a single byte of source color data does not contain color information for more than one source pixel, decision box 100 determines “No” and one byte of source color information for each source pixel. If so, decision box 100 determines No, and the byte value of the unconverted source color data for the next source pixel in the source row is the unconverted source pixel value for the last converted source pixel. The color data is compared to determine if they are the same. If these values are the same, then the source pixel is the same color. This color data is compared byte by byte if the format is single byte,
In the case of a 3 byte, or other multibyte format, all multibytes are compared.

In the decision box 100, the decision is "Yes".
, Each source byte of color data contains color information for more than one source pixel. In this case, a second test is performed at decision box 102 to determine if all pixels in the byte are the same color. This is done by comparing the color data values for each source pixel in the byte. If the color data values are the same, then the source pixels are the same color. If the decision is "Yes", the source pixels are of the same color and processing proceeds to compare the source bytes as if this data were contained in a 1-byte source byte.

If the decision in decision box 102 is "No", then the source pixel is a different color and conversion to grayscale values and dithering continues as described above for box 74.

If the dither pattern widths of the output position bits generated from the conversion of the source pixels are all the same grayscale value and it is determined that the unconverted color data are comparable, the dither processing is performed. Source address and last translated source
The source color value of the pixel is stored in storage boxes 106 and 104 as shown in FIG. When a save occurs, this process looks ahead at the source pixel row to determine if the source color values stored in save box 104 are the same as each new source pixel color data value. First, decision box 108 checks the new color data values. If they are the same, the address of the source pixel examined in decision box 108 is updated in turn with the address of the last source pixel examined in decision box 108 to be "same" and saved in save box 110. Stored in. This process
It continues until decision box 112 determines that the end of the source image line has been reached, or until the decision box 108 determines that the source pixels are "not identical."

If the test performed in decision box 108 indicates "non-identical" source color data, or decision box 112 detects that the end of the line has been reached, it is saved in save box 106. The check box to determine if the address of the last converted source pixel is the same as the address of the source pixel that was last checked for color data and saved in save box 110. Done at 114. Source·
If the addresses of the pixels are the same, no lookahead of the source image line is done. If the source pixel address is not the same, then each source pixel of the same color as the first converted source pixel in this operation cycle is sourced without conversion and dithering.
Read ahead image lines.

Next, the process determines the total number of scaled output pixels required to produce the scaled process from the beginning of the output image row to the same relative position in the destination image, shown in box 118. In order to determine the,
In box 116, the number of source pixels from the beginning of the source image row to the address of the last stored source pixel in box 110 is calculated. box
At 120, the total number of output pixels generated from the read-ahead skip can be calculated by subtracting the number of output pixels already sent to the output buffer from this total number.

It is preferable to calculate the number of output pixels generated from the source pixels thus skipped by the look-ahead. Because the scaling process may scale the source pixels to the output pixels at a non-integer rate, so that the cumulative sum of the fractional output pixels reaches the full output pixel every time. This is because various output pixels are added periodically.
Also, by calculating the number of output pixels in this way, a check is made in decision box 72 to determine if all of the output position bits are located in the bit mask element, i.e. if they are the same color. Previously, at the beginning of the cycle where only the dither pattern width of the output position bits is accumulated in the bit mask array, we eliminate any scaling issues that may occur when performing the initial conversion and scaling process. If all of the output position bits are not necessarily the same bit mask element, the dither pattern widths of the output position bits are sequentially dithered to produce additional scaled output bits from the last read source pixel. All carryovers are added back as the first source pixel grayscale value and output position bits in the next conversion cycle. If the result of the check in decision box 72 is "Yes", then these remaining same output position bits are the output bits needed to reach the same relative position in the output image during the calculation in box 118. It is taken into account in the calculation of the total number.

After the total number of output bits of the same color has been determined, the grayscale value from the grayscale value array element 200 is selected in box 74 to determine the grayscale value and the output image dither pattern tile in the tile. Position Y _N
As shown in box 76 for the exact row Y _DN of the exact dither pattern matrix, as well as the dithering for the output pixels generated from the first bit mask array element 201 in a cycle without read ahead. Used to choose. Similarly, the first dither matrix bit X in row Y _DN
The selected dither pattern row is rotated in box 80 so that the _DN corresponds to the same position X _N of the first output position bit of bit mask element 201 when printed in the dither tile of the output image. It The bit mask element 201 is then boxed at box 82 to ensure that the full dither pattern tile width is produced by ANDing with the rotated rows of the dither pattern matrix to produce the actual output pixels. Is ORed to.

If no compression is done, box 96 in FIG.
The raster rows are created in the output buffer for final transfer to the print engine, as shown in.

If data compression is to be done, then at this point the calculation of the number of output bytes is performed in box 86, as shown in FIG. 6, by the number of output bits from box 120 and the dither pattern from bit mask array element 201. It is done by dividing the width plus one by eight. Lookahead and scaling ensures the final number of output bytes, and may result in insufficient output bits remaining to produce the final full byte. This remainder is stored in storage box 122, along with the grayscale value in box 54, for use in the next conversion cycle, until the end of the line is reached, where the final partial output byte is created.

At this time, the first complete dither pattern width of the output bit is divided into two repeating bits of output data. It is also known at this point how many repeated bytes are needed in the output string of data produced in this cycle. The dither pattern matrix used in the present invention is determined empirically, and if possible, it is likely that the row will have the same value even if it is rotated when it is split into output bytes. 92
For a compressed data string entry as shown in Figure 3, each row of the matrix has a symmetric pattern so that its output is created as a repeating pair of output bytes.

Whether or not the compressed data string is created in box 92, or 94, it is sent to the output buffer as shown in box 96, where the full compressed data command for the target image raster row is sent. , Created for transfer to the final printer, or other device via output port 98.

Simultaneous with the transfer of the repeat run compressed data entry in box 92, a command is sent in box 124 to zero the grayscale and bit mask arrays. The only exception is if the box 122 has a remainder, the grayscale value for the remaining bits in the save box 122 is the same as the value of the element 200 in the box 54, so this element is not reset to zero. Instead, the remaining bits are added back as the first output position bit entry of the bit mask element 201, as shown in box 56.

Scale in the X direction when a complete raster row of output image pixels is assembled in the output buffer for transfer to a printer, or other device, whether or not the conversion process includes compressed data. All of the pixelized pixels are calculated, but the rest in the Y direction is not scaled. For this reason, the same row of source image pixels is reused with sufficient time to scale in the Y direction, eight rows in this example, but one additional row with one cumulative scaling fraction. When, the ninth row is included cyclically. This process is the same, but it does not necessarily result in the same output raster row because the position Y _N in the output image tile changes, but different dither
It may or may not contain a pattern, resulting in a different row Y _DN being selected in the dither matrix.

Assume that the processor 10 is incorporated into the printer as to the host processor, or that the data produced by the scaled conversion process is incorporated with respect to any other device that is used directly rather than by transfer. In the arrangements shown in FIGS. 2 to 6, data compression as described above with reference to FIG. 2 is not necessary. In this case, the scaling process is shown in the overall format of the configuration of FIG. 2 used in FIGS. 2, 3, 4, and 5. In this configuration, the first scaled row Y _N is created and stored in the output buffer in exactly the same manner as in the conventional configuration where data compression was used, as shown in box 96 of FIG. , To the printer engine. However, after this the process is slightly modified to take advantage of the repeating row Y _DN in the dither pattern matrix. These repeating rows Y _DN in the dither pattern matrix are available in a variety of situations, including, but not limited to, box 48 in FIG.
This includes situations where Y _S in the scaling factor exceeds the number of rows Y _DN of the dither pattern matrix. This is the next 2
It can occur in either of two situations. One is row Y where the dither pattern matrix repeats in the dither pattern tile.
_In situations where the number of _DNs is constrained, for example even if a 16-row dither pattern matrix is used, the DN
Is in the range of 1 to 4, and the other is when the scaling factor Y _S exceeds the number of rows Y _DN in the dither pattern, for example Y _S = 32 rows and the dither pattern is Contains 16 rows, in which case two dither pattern tiles are printed in the output image for each source image pixel. Therefore, using this repeating row feature,
Once a raster row Y _N has been created in the output buffer memory, as shown in, a decision box 128 determines if there are more than Y rows to create from the source row in the scaling process. Yes, that is, Y _N is less than or equal to Y _S. The judgment in decision box 128 is “Y
es ”, a decision box 130 determines if the next row Y _{N + 1} to be generated in the scaling process is greater than the last or last row in the non-repetitive dither pattern matrix Y _DNL . Is done. If the next row Y _{N + 1} is greater, the next row is a repeat of the row already created in the output buffer, and in box 132 row Y _{N + 1-DNL} is selected from the output buffer and Sent to the printer engine via 98.

If the row Y _N created in box 96 was the last row in the scaling process as determined in decision box 128, then the elements in both the grayscale and bit mask arrays in box 124 are Zeroed and the process is restarted by adding the scaled output position bits to the first element of the bit mask array, as shown in box 56. If the next row Y _{N + 1} is less than the dither pattern repeat value Y _DNL , as determined by decision box 130, then the scaling process is correct for the dither pattern, as shown in box 76 of FIG. Continue selecting the next position Y _DN in the matrix. Thus, if there are repeating rows of data that use repeating rows of the dither pattern matrix, those rows stored in the output buffer are reset rather than recreated.

Although this example includes scaling up from low resolution to high resolution, in practice a similar process can be used for scaling down from high resolution to low resolution. For example, there is a scale down from a 600 DPI rasterized source image to a video monochrome image. In such a case, the scaling process is reversed and only selected source image pixels are examined and included in the process. For example, from a 600 DPI rasterized source color image to a 300 DPI
If a scaling process to a monochrome image is performed, every other source pixel is converted in the conversion cycle and / or read in a repeated data look-ahead cycle. Similarly, this process is only used in situations where image clipping occurs, by simply correcting the identification of the beginning of each source line and the end of line inspection. Similarly, this process can be used to up-size or down-size the scaling process. Finally, this process can be used to read and scale monochrome source images.

While the preferred embodiments of the present invention have been shown and described, the invention is not limited thereto but can be variously practiced to practice within the scope of the claims of the invention. Clearly understood.

The embodiments of the present invention are listed below.

1. A processor unit for converting source pixel image data into rasterized monochrome bitmap rows, the processor unit including a random access memory, a processor, and a control memory, the control unit comprising:・ The memory is a grayscale value conversion table,
Multiple dither matrices, a procedure for determining the position of an output pixel in an output image dither tile for a rasterized monochrome bitmap, factored and scaled output in both X and Y directions A processor unit characterized by having a scaling transformation procedure for generating position bits, comprising: a. Establishing in said random access memory an array of elements for storing grayscale values. B. Establishing an array of bit mask elements in the random access memory, each bit mask element having a bit block for storing output position bits, Each of which is indexed to the corresponding grayscale value element, and each of the bit mask elements is output image Sized to hold multiple output position bits equal to the number of bits in the dither pattern tile row, c. Convert the source pixel data information to the corresponding grayscale value, and d. Generating scaled output position bits for the pixels, e. Storing each different grayscale value resulting from the transformation of the data information in a separate element of the grayscale value array, f. Storing the output position bit in an element of a bit mask array indexed to an element of a grayscale array for the source pixel being generated, the output position bit being associated with another source pixel after conversion; Stored in a bit mask element at the same relative position as the pixel, relative to other output position bits , G. Equal to the number of bits output image dither pattern tile row, until a sufficient number of source pixels to generate a plurality of output position bits are converted, step (c)
Repeat steps (f) to (h) and perform a check to determine if all of the output position bits have been stored in one bit mask element, i. bit·
If indicated to be stored in the mask element, the unconverted source pixel data information and the last for the next source pixel until the end of the source line or data information with a different value appears. Comparing the unconverted data information for the converted source pixels of, j. Storing the address of the last source pixel with a value of data information equal to the data information of the last converted source pixel, k. Calculated the number of output position bits produced from the inspected and counted source pixels having the same value as the data information and stored in the bit mask element to produce L. bits of output data Dithered the output position bits and m. Examination shows that the output position bits are stored in more than one bit mask element. A source image, wherein the output position bits stored in each element of the bit mask array holding the stored output position bits are dithered to produce bits of the output data. A method of converting data into output data for a rasterized monochrome bitmap.

2. n. Compress the bit output data for the number of output position bits calculated in step (k) into a compressed data output signal, o. Assemble the output signal from the bits of the output data before assembly from step (m) And forming an output signal.

3. To generate a number of additional rows of rasterized monochrome bitmaps equal to the scaling factor Y, steps (c) to (m) with respect to the source pixels transformed in steps (c) to (g). ) Is further characterized by repeating No. 1
The method described in.

4. An additional step (q) that counts the number of bits of output data stored in the output buffer for each row of the rasterized monochrome bitmap.
The method according to item 1, further comprising:

5. Step (k) runs from the beginning of the k (1) source line to the last transformed source
Compute the number of source pixels up to the last compared source pixel with data value equal to the pixel, and k (2) monochrome bitmap row for the source pixel number calculated in step k (1). Calculate the number of scaled output position bits to be generated from the beginning of k (3), from the number calculated in step k (2), the number of bits of output data for the row that was previously generated and stored. The method according to Item No. 1 or No. 4, further characterized by:

6. Step (L) calculates the position Y in the dither pattern tile for the output position bit of the first position in the L (1) filled bit mask element, and L (2) filled bit mask To calculate the position X in the dither pattern tile for the output position bit of the first position in the element and to select the L (3) dither pattern matrix, of the bit mask array filled with output position bits Rotate the selected dither pattern matrix to relocate L (4) to the first position in the dither pattern matrix using the grayscale value stored in the element corresponding to the element, Bits of dither pattern matrix are step L
(1) and L (5) have positions corresponding to position X and position Y calculated in L (2), to generate bits of L (5) output data.
6. The method according to item number 1, 4, or 5, further characterized by ANDing the output position bits of the mask element and the rotated dither pattern matrix.

7. Step (n) calculates the number of output bytes by dividing the number of output position bits calculated in n (1) step (k) by 8, and n (2) each of the output position bits into an 8-bit byte. Grouping, n (3) saving the remaining output position bits, grouping the bits of the output data from the n (4) step (L) into an output signal byte, n (5) generated in step n (4) Comparing the output signal bytes to determine if the output signal bytes are the same or different, and that the output signal bytes are different due to the comparison made in n (6) step n (5). , The output signal byte generated in step n (4), and the step
The number of output bytes determined in n (1) is used to create a character run of output signal bytes, and the output signal bytes are said to be the same by the comparison made in n (7) step n (5). , A bit mask is created to create an iterative run-compressed data output signal and include n (8) output position bits equal to the rest of the output position bits determined in step n (3). Item 3. The method according to item 2, further characterized by rewriting the first element of the array.

8. Step (m) ORs the set of bits of the output data before assembly with the bit of the output data generated in the check array generated in step (n) of m (1) and outputs m (2) 8. A method according to item number 3, 4, 5, 6 or 7, further characterized by dividing the set of bits of the output data into 8 groups to form a signal byte.

9. Step (i) performs a check to determine if a byte of i (1) source image data contains information about more than one source pixel, i (2) step i (1) Sources by inspection carried out in
More than one source for each byte of image data
If it is shown that there is data for a pixel, then the data for each of the source pixels contained in each byte is examined to determine if they are equal, i (3) step i According to the inspection conducted in (2),
If the data for each of the source pixels contained within a byte of image data is shown to be the same, until the end of the source line is reached or until a byte of data information with a different value appears, By comparing the unconverted bytes of the unconverted data information for the source pixel of with the unconverted bytes of the data information for the last converted source pixel, i (4) by the check done in step i (2) Source·
If the data for each of the source pixels contained within the bytes of the image data is shown to be different, skip steps (j) through (m) and skip i (5) step i (1 ), The data for more than one source pixel is not contained within each byte of the source image data, until the end of the source line is reached, or the value Further comprising comparing the unconverted byte of unconverted data information for the next source pixel with the unconverted byte of data information for the last converted source pixel until another byte of different data information of Yes, item number 1
The method described in.

10. A processor unit for converting source pixel image data into rasterized monochrome bitmap rows, the processor unit including a random access memory, a processor, and a control memory, the control unit comprising:・ The memory is a grayscale value conversion table,
Multiple dither matrices, a procedure for determining the position of an output pixel in an output image dither tile for a rasterized monochrome bitmap, factored and scaled output in both X and Y directions Scaled conversion procedure for generating position bits, storage for storing rows of rasterized monochrome bitmap, and source of output data bits in row X of rasterized monochrome bitmap A processor having a method for converting image data
A unit, the method comprising: a. In the random access memory, establishing an array element for storing grayscale values; b. In the random access memory, defining a bit mask element. Establishing, each bit mask element having a block of bits storing an output position bit, each of said bit mask elements being indexed to a corresponding grayscale value element, the output image dither pattern
A scaled output position bit size in the X direction to hold multiple output position bits equal to the number of bits in the tile row, and a raster generated for each source pixel to be transformed. Compute the number of scaled rows Y of the digitized bitmap, and d. Determine the number of rows Y generated before the row Y with repetitive data appears in the selected dither pattern matrix. , E. Transforming the data information of the source pixel into the corresponding grayscale value, f. Producing scaled position bits for the transformed source pixel, g. Said in individual elements of the grayscale value array, Store each different grayscale value resulting from the transformation of the data information, and h. Output position bits into the grayscale array element for the source pixel that is generated. Storing the output position bits in a dexed bit mask array element, the output position bits in the bit mask element in the same relative position with respect to the other output position bits, the other transformed source Stored as the relative position of the source pixel with respect to the pixel, i. The source pixel is transformed enough to produce multiple output position bits equal to the number of bits in the output image dither pattern tile row. Repeat steps (e) through (h) until j. Check to determine if all of the output position bits are stored in one bit mask element, and k. All one bit
If it is shown stored in the mask element, the unconverted source pixel data information and the last conversion for the next source pixel until the end of the source line is reached or different values of data information appear. Compare the unconverted data information for the source pixel that has been converted, L. store the address of the last source pixel that has a value of data information equal to the data information of the last converted source pixel, m. Output position bits stored in the bit mask element to calculate the number of output position bits generated from the examined and counted source pixels having the same value of n. Dither to, o. If the output indicates that the output position bits have been stored in more than one bit mask element, the output Dither the output position bits stored in each element of the bit mask array, which holds the stored output position bits to produce the bits of the data, and p. Storing, q. Assembling the rasterized bitmap row in storage, and r. Generating an additional row number of the rasterized monochrome bitmap before row Y with repeated data appears. Step (e) or step with respect to the source pixel transformed in step (e) or step (i)
Repeat (q), and s. Rasterized bitmap if additional scaled row Y of rasterized bitmap with repeating data is generated from scaling the first source pixel. A repeated row Y is selected from the stored rows and retransmitted.

[0072]

In accordance with the present invention, a process for a string of source pixels that has a repeating portion, which need not all be converted to grayscale values, and which scales from the scaling of the source pixels. All of the output pixels are provided that need not be individually dithered to produce an all-rasterized bitmap.

[Brief description of drawings]

1 is a processor adapted to carry out the present invention;
FIG. 3 is a high level block diagram of the unit.

FIG. 2 is an overall flow diagram showing the overall procedure of the present invention without any compression of the output image signal.

FIG. 3 is an overall flow diagram showing the overall procedure of the present invention without any compression of the output image signal.

FIG. 4 is an overall flow diagram showing the overall procedure of the present invention without any compression of the output image signal.

FIG. 5 is an overall flow diagram showing the overall procedure of the present invention without any compression of the output image signal.

FIG. 6 is an alternative to FIG. 5, which is related to FIGS. 2-4 and represents an overall flow diagram illustrating the overall procedure of the present invention using an output data compression procedure.

FIG. 7 is a diagram showing a conversion process used according to the procedure of FIGS. 2 to 6;

FIG. 8 is a diagram showing a conversion process used according to the procedure of FIGS. 2 to 6;

[Explanation of symbols]

 10 Processor unit 12 Microprocessor 14 Random access memory (RAM) 16 Storage memory 18 Input port 20 Output port 22 Bus

Claims (1)

[Claims] 1. A processor unit for converting source pixel image data into rasterized monochrome bitmap rows, the processor unit including a random access memory, a processor, and a control memory. Where the control memory determines a position of an output pixel in an output image dither tile for a grayscale value conversion table, a plurality of dither matrices, a rasterized monochrome bitmap, X, And a scaling unit for producing scaled output position bits in both the Y and Y directions, the processor unit comprising: a. In the random access memory; , Element allocation for storing grayscale values B. Establishing an array of bit mask elements in the random access memory, each of the bit mask elements having a bit block for storing output position bits; Each mask element being indexed to a corresponding grayscale value element, each bit mask element holding a plurality of output position bits equal to the number of bits in the output image dither pattern tile row; The size is determined, c. The source pixel's data information is converted to the corresponding grayscale value, d. The scaled output position bits for the converted source pixel are generated, and e. The grayscale value array's Store each different grayscale value resulting from the transformation of the data information in a separate element and generate the output position bits f. Storing the output position bit in an element of a bit mask array indexed to an element of the grayscale array for the source pixel to be converted, the output position bit of the source pixel being associated with another converted source pixel. Generate multiple output position bits, stored in a bit mask element at the same relative position as the relative position relative to other output position bits, g. Equal to the number of bits in the output image dither pattern tile row Step (c) until enough source pixels have been converted to
Repeat steps (f) to (h) and perform a check to determine if all of the output position bits have been stored in one bit mask element, i. bit·
If indicated to be stored in the mask element, the unconverted source pixel data information and the last for the next source pixel until the end of the source line or data information with a different value appears. Comparing the unconverted data information for the converted source pixels of, j. Storing the address of the last source pixel with a value of data information equal to the data information of the last converted source pixel, k. Calculated the number of output position bits produced from the inspected and counted source pixels having the same value as the data information and stored in the bit mask element to produce L. bits of output data Dithered the output position bits and m. Examination shows that the output position bits are stored in more than one bit mask element. A source image, wherein the output position bits stored in each element of the bit mask array holding the stored output position bits are dithered to produce bits of the output data. A method of converting data into output data for a rasterized monochrome bitmap.