JPH04217172A - Picture processor - Google Patents

Abstract

PURPOSE:To realize requantization capable of reserving the density and to obtain a device accurately correcting the recording irregularity for each recording element by executing the requantization for each picture element forming the final output based on the peripheral requantization data. CONSTITUTION:An irregularity correction ROM 34 stores the irregularity data correcting the recording density irregularity generating addresses corresponding to each recording element, and the irregularity data on the periphery of the noted picture element is successively read out by the address circuit not shown in the figure. F/F 35-1 to 35-4 held the delay of the irregularity data for each picture element, and the output of the ROM 34 corrects binary data (a) and (f), the output of the F/F 35-1, (b) and (g), the output of the F/F 35-2, (e) and (h), F/F 35-3, (d), (i), and (k), F/F 35-4, (e), (j), and (l). Multiple devices 30-a to 30-k performs the multiplication between correction data and the binary data located in (a) to (k), and the weight coefficient according to the location is multiplied. An adder 32 adds 12 elements of multiplication result to obtain the weighted average by multiplying it 1/63 times through removing device 33.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention is applicable to digital copying machines, FAX machines, etc.
, regarding an image processing device having a recording device such as a printer,
In particular, the present invention relates to an image processing apparatus that corrects recording unevenness for each recording pixel.

0002

2. Description of the Related Art Conventionally, an error diffusion method has been known as one of the pseudo halftone processing methods. Furthermore, devices that arrange solid-state recording elements and record in line sequential order include LCD printers and LED printers.
Many printers, inkjet recording systems, etc. are known, but since each recording element cannot record uniform dots, it is known to correct the unevenness of recording elements for each recorded image data.

[0003] Among these methods, the error diffusion method is effective because it can almost preserve the gradation and resolution information of the input image, even though the input image data is requantized to a smaller number of levels. By correcting the recording element unevenness of the input data before requantizing the recording signal using the method, it becomes possible to record while suppressing the occurrence of unevenness.

0004

[Problem to be Solved by the Invention] However, in the error diffusion method, the error that occurs during requantization is diffused in the processing direction to the next pixel on the same line and the same pixel on the next line, so it is difficult to maintain fidelity to the recorded pixels. It is not possible to correct unevenness.

[0005]

[Means for Solving the Problems] The present invention has been made for the purpose of solving the above-mentioned problems, and has the following configuration as a means for solving the above-mentioned problems. That is, a first correction means corrects requantized pixel data of interest with correction data corresponding to the position of the pixel of interest, and a plurality of pieces of data near the pixel of interest corrected by the first correction means a threshold calculation means for calculating a threshold value for requantizing pixel data; a quantization means for requantizing pixel data of interest using the threshold value; a second correction means for correcting an error caused by the requantization; It is equipped with an output means for outputting converted data.

[0006] The threshold value calculation means calculates a threshold value from a plurality of pieces of data in the vicinity excluding the pixel of interest. The threshold value calculation means calculates a threshold value from a plurality of pieces of neighboring data including a predicted value of the pixel of interest.

[0007]

[Operation] In the above configuration, since recording unevenness is taken into account and requantization is performed for each pixel that forms the final output based on requantized data in its vicinity, more accurate recording unevenness correction is possible. Become.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings. [First Embodiment] FIGS. 1 and 2 are block diagrams of an embodiment according to the present invention. In FIGS. 1 and 2, original image information to be processed is read by an image signal input section 1-1 such as a scanner, and quantized into 8-bit 256-gradation by a quantization section 1-2. 3 is an unevenness correction average value calculation unit that calculates a weighted average value after correcting unevenness of binary data for 12 pixels;
4 is a latch in which quantized image data is set, and 256 is 8 bits in the case of gradation. 5 is a subtracter, 6 is a 1/2 divider, and the output e1 is added to the pixel next to the pixel of interest. e2 is added via the line memory 7 to the same pixel delayed by one line.

Further, 8 compares the binarized average value obtained by the unevenness correction average value calculating section 3 with the input image data, and binarizes the input image data depending on whether it is larger or smaller than the binarized average value. The output from the comparator 8 is output to the recording device 20, and is also input to a line memory 2-f and a flip-flop (hereinafter referred to as "F/F") 2-k. The line memory 2-f delays the binary data by approximately one line and outputs it to the F/F 2-g.
input to a, delay one more line, and switch to F/F2
- Output to b.

[0010] The output of F/F2-k is transferred to F/F2-l,
/F2-g output to F/F2-h, F/F2-i, F/
Connect the output of F/F2-b to F2-j, F/F2-c, F/
If F2-d and F/F2-e are delayed by the pixel clock and held, the binary data a to k obtained at each F/F input/output terminal indicates multiple pieces of data near the pixel of interest. As shown in Figure 3, the pixel position of interest (
*) Located adjacent to.

[0011] That is, k is the data that was binarized immediately before,
h is binary data one line before the pixel position of interest. Note that the above-mentioned unevenness correction average value calculation unit 3 calculates a weighted average value after correcting unevenness of the binary data for the 12 pixels. Now, the subtracter 5 calculates the difference between the input image data after error correction and the weighted average value obtained from the binary data of the adjacent 12 pixels after unevenness correction, and uses the output as a binarization error. This error is divided into two by the above-mentioned distributor 6, and one output (e2) is held in the line memory 7 with a delay of one line. The other main power e1, together with the error data e2 of the previous line outputted from the line memory 7, corrects the error of the next pixel input image data in the adder 9. The error-corrected image data is input to the latch 4, and when the next pixel clock is input, it is read out to the subtracter 5 and comparator 8, and the above-described error correction and binarization operations are performed sequentially.

Next, details of the unevenness correction average value calculating section 3, which is a characteristic configuration of this embodiment, will be explained with reference to FIG. In FIG. 4, the unevenness correction ROM 34 stores 6-bit wide unevenness data for correcting recording density unevenness that occurs at addresses corresponding to each recording element, and is sequentially stored in the vicinity of the pixel position of interest by an address circuit (not shown). The unevenness data is read out.

[0013] F/F35-1, 35-2, 35-3, 3
5-4 is an F/F for delaying and holding read unevenness data for each pixel; the output of the ROM 34 is binary data a and f, and the output of the F/F 35-1 is b, g, and F. /F35
-2 outputs are c and h, F/F35-3 outputs are d, i,
The output of k, F/F 35-4 corrects the binary data of e, j, and l.

Multipliers 30-a, 30-b,..., 30-k
perform multiplication of the correction data and the binary data located at a to k, respectively, and then multipliers 31-a to 31-3 respectively.
1-k is used to multiply by a weighting coefficient according to the position shown in FIGS. 6 and 7. The adder 32 converts the above multiplication result into 12
This is an adder that adds pixels, and the addition result is sent to divider 3.
If you multiply by 3 to 1/63, the input image data width (0 to 255
, 8 bits) is determined. Note that the total sum of weights shown in the figure is set to 255.

As explained above, according to this embodiment, after correcting the unevenness by associating the binary data with the recording element, the unevenness is corrected in the vicinity of the position of interest in order to obtain the weighted average value and use it as the binarization threshold. If only faint dots can be recorded due to this, the average value will be correspondingly smaller, making it easier to generate a recording "1" signal, and the binarization error will be calculated from a weighted average value that takes this unevenness into account. In order to correct errors, it becomes possible to perform binarization while correcting unevenness more accurately. Particularly when applied to a copying machine, it is efficient because the image binarization section and unevenness correction can be performed at the same time.

[Second Embodiment] Next, a second embodiment of the present invention will be explained in detail with reference to FIGS. 6 and 7. In FIGS. 6 and 7, the same reference numerals as those in FIGS. 1 and 2 are given to the same components, and detailed description thereof will be omitted. The second embodiment shown in FIGS. 6 and 7 is a weighted average value m1 when the unevenness correction average value calculation unit 30 predicts the binarization of the pixel of interest and predicts that the binarization of the pixel of interest is recorded as "1". and the weighted average value m0 when the binarization of the pixel of interest is predicted to be non-recording “0”
is calculated, and using the adder 12 and the divider 13, (m0+
The comparator 8 binarizes m1)/2 as the binarization threshold, and
If the selector 11 is “1” according to the value conversion result, m1 is
In the case of "0", m0 is input to the subtracter 5 to obtain the binarization error. In other words, in the case of recording "1", the difference from m1 is the binarization error, and in the case of non-recording "0", the difference from m0 is 2
This results in a valuation error. Therefore, the binarization process is more accurate than that of the embodiment described above.

Next, details of the unevenness correction average value calculating section 30 in the second embodiment shown in FIGS. 6 and 7 will be explained below with reference to FIG. The difference between the calculation unit 30 shown in FIG. 8 and the unevenness correction calculation unit 3 shown in FIG. The point is that the total sum is set to 255. Therefore, the value obtained by multiplying the weighted output of the adder 32 by 1/63 after correcting the unevenness of the binary data of a to k is the average value m0 that predicts the pixel of interest to be binarized to "0". , the output value of the adder 32 is multiplied by the weight "38" of the pixel of interest and the unevenness data of the pixel of interest position by the multiplier 30-x, and the value added by the adder 36 is multiplied by 1/63 to obtain the pixel of interest. “1
” is the average value m1 that is predicted to be binarized.

Note that the unevenness correction ROM shown in FIGS. 4 and 8 includes:
8-bit width data is stored, and dividers 33-1, 33-
By setting 2 to 1/255, more accurate correction is possible. On the other hand, when the fluctuation range of unevenness is small,
Instead of storing data pixel by pixel, if data is stored every other pixel and two consecutive pixels are corrected using the same correction data, it is possible to perform the correction using a cheaper and smaller capacity ROM.

Furthermore, the stored data is not limited to semi-fixed data, but can be measured as time-varying data using a RAM, and more accurate correction can be made by sequentially updating the data. As explained above, according to the present embodiment, similarly to the first embodiment, after correlating binary data with the recording element and correcting unevenness, the weighted average value is obtained and used as the binarization threshold. If only faint dots can be recorded near the target position due to unevenness, the average value will be correspondingly smaller, making it easier to generate a recording "1" signal, and binary conversion will be performed using a weighted average value that takes this unevenness into account. Since the error is calculated and corrected, it is possible to perform binarization while correcting unevenness more accurately. Particularly when applied to a copying machine, it is efficient because the image binarization section and unevenness correction can be performed at the same time.

[Third Example] Next, using FIGS. 10 and 11, a third example according to the present invention will be explained.
Examples will be explained in detail. The embodiment shown in FIGS. 10 and 11 differs from that shown in FIGS. 1 and 2 in that the input data to be requantized is 1 bit. In other words, this is another embodiment in which data that has been binarized as recording data is corrected for unevenness and then binarized again without considering the unevenness characteristics specific to the recording apparatus.

In FIGS. 10 and 11, input 1-bit data is input to the selection input terminal of the selector 10, and "255" is sent to the adder 9 for recording "1" and "0" for non-recording. Output it as image data. Therefore, the subsequent operation is similar to that of the first embodiment, and re-binarization is possible while correcting the element unevenness inherent in the recording element to which the binary data is input.

[0022] Compared to the first and second embodiments, this embodiment has the following points:
It can be configured as a standalone printer.

[Other Embodiments] The first to third embodiments described above are all examples in which images are binarized, but even when performing multilevel pseudo halftone processing at two or more levels or several levels, Needless to say, it can be implemented in the same way. Furthermore, although the embodiment has been described as an example of line-sequential raster recording, in the case of a recording device or a display device that has a group of recording elements in a planar manner, two-dimensional correction can be performed by using two-dimensional correction data. It is also possible to record and display the data at the same time.

As explained above, according to this embodiment, a scanner for reading an image, for example, is stored in the read-only memory.
Alternatively, if correction values based on the recording characteristics of the recording side element are stored, accurate images can be reproduced. Further, if a rewritable RAM is provided with correction values as data on changes over time, changes over time in the scanner and recording element can be corrected.

[0025]

As described above, according to the present invention, it is possible to provide an image processing apparatus that realizes requantization that can preserve density and more accurately corrects recording unevenness for each recording element.

[Brief explanation of the drawing]

[Figure 1] and

FIG. 2 is a block diagram showing the configuration of a first embodiment according to the present invention;

FIG. 3 is a diagram showing a plurality of pieces of data near the pixel of interest in this embodiment;

FIG. 4 is a block diagram showing the detailed configuration of the unevenness correction average value calculation unit in FIGS. 1 and 2;

FIG. 5 is a diagram showing weighting values near the pixel of interest in FIG. 4;

[Figure 6] and

FIG. 7 is a block diagram showing the configuration of a second embodiment according to the present invention;

FIG. 8 is a block diagram showing the detailed configuration of the unevenness correction average value calculation unit in FIGS. 6 and 7;

9 is a diagram showing weighting values near the pixel of interest in FIG. 8,

[Figure 10] and

FIG. 11 is a block diagram showing the configuration of a third embodiment according to the present invention.

[Explanation of symbols]

1-1 Image signal input section 1-2 Quantization section 2-2, 2-f, 7 Line memory 3, 30
Unevenness correction average value calculation section 5 Subtractor 6 1/2 divider 8 Comparator 20 Recording device 34 Unevenness correction ROM

Claims (3)

[Claims] 1. A first correction means for correcting requantized pixel data of interest with correction data corresponding to the position of the pixel of interest; a threshold calculation means for calculating a threshold value for requantizing the pixel data of interest from the data; a quantization means for requantizing the pixel data of interest using the threshold; and a second correction means for correcting an error caused by the requantization. , an image processing device comprising an output means for outputting requantized data. 2. The image processing apparatus according to claim 1, wherein the threshold value calculation means calculates the threshold value from a plurality of pieces of data in the vicinity excluding the pixel of interest. 3. The image processing apparatus according to claim 1, wherein the threshold value calculation means calculates the threshold value from a plurality of pieces of neighboring data including a predicted value of the pixel of interest.