JPH0955855A - Picture processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a picture processor where a pseudo half-tone pattern is sufficiently eliminated and blur of a picture is suppressed to satisfactorily restore the gradation at the time of estimating original gradation information from picture data where the gradation is limited. SOLUTION: Picture data is inputted from a picture input part 1 and is stored in a picture storage part 2. A picture element value integrating part 11 reads in a picture of a prescribed area including a given picture element from the picture storage part 2 and integrates picture element values of the given picture element and plural apertures set around it. A weighting value setting part 12 sets a weighting value corresponding to each of plural integrated values obtained by the picture element value integrating part 11. An average value calculation part 13 performs the weighted mean processing in accordance with integrated values obtained by the picture element value integrating part 11 and weighting values set by the weighting value setting part 12 to obtain the restoration value of the given picture element. The obtained restoration value is sent to a picture output part 3 in order and is outputted.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for processing a digitized image, and more particularly to an image processing apparatus for estimating original gradation information from an image subjected to limited gradation processing and restoring the image. Is.

[0002]

In order to reduce the capacity of the storage device and the transmission time when storing / transmitting a digital halftone image, 256
An image expressed in multiple gradations such as values may be converted into a small number of gradations such as binary or quaternary by a pseudo halftone process such as a dither method or an error diffusion method. However, such a pseudo-halftone image may generate a pattern called moire, which is not included in the original image, when the image is enlarged or reduced. Further, when using an output device capable of multi-value recording at the time of output, there is a problem that the performance of the output device cannot be utilized. Therefore, as one of means for solving these problems, a method of restoring the original gradation information from the pseudo halftone image has been studied.

Conventionally, as a method for restoring an original image from an image subjected to limited gradation processing, for example, Japanese Patent Publication No. 5-1146.
As described in Japanese Patent Laid-Open No. 5 and the like, it is most common to use a fixed size aperture to smooth an image.
However, in such a method, if the opening is large, even a portion where density changes such as an edge is smoothed uniformly,
The image is blurred. Further, if the aperture is made small, a pseudo halftone pattern remains, or the number of gradations after restoration becomes insufficient. For example, even if a binary image is processed with an opening having a size of 3 × 3, it can be restored to only 10 values.

Therefore, as disclosed in Japanese Patent Publication No. 6-24006, Japanese Patent Publication No. 4-31466, Japanese Patent Publication No. 2-76370, etc., a plurality of apertures are prepared and the apertures are opened under various conditions. As described in the selection method, IEICE Technical Research Report, IE93-13, “Estimation of Halftone Image from Pseudo Halftone Image”, etc., a regression line is obtained from the average value of the neighborhood area and restored. The method of doing is proposed.

However, in the method of selecting a plurality of apertures, a small aperture is not selected and the pseudo-halftone pattern is not removed due to an edge or fluctuation of the pseudo-halftone pattern. The image may become unnatural by changing to. Further, since the smoothing is basically performed by the opening, there is a problem that the number of gradations after restoration cannot be sufficiently obtained. In addition, the method using a regression line is difficult to follow a rapid density change because it is an estimation by a straight line, and the restoration values are separately obtained from the regression lines in the 8 or 4 directions around the pixel of interest, and the average is taken. Blurring may occur due to averaging.

[0006]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems of the existing method, and when estimating the original gradation information from the limited gradation-converted image data, the pseudo halftone is used. An object of the present invention is to provide an image processing device capable of sufficiently removing a pattern and capable of restoring good gradation while suppressing blurring of an image.

[0007]

According to a first aspect of the present invention, in an image processing apparatus, a pixel value integrating means for setting a plurality of openings around a target pixel and a pixel value for each opening is provided. A weight value setting means for obtaining a weight value for each aperture from the integrated value obtained by the pixel value integrating means, an integrated value obtained by the pixel value integrating means, and a weight value obtained by the weight value setting means An average value calculating means for calculating a weighted average value is provided, and gradation restoration processing is performed by the pixel value integrating means, the weight value setting means, and the average value calculating means with each pixel in the given image as the target pixel. It is characterized by that.

According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the weight value setting means selects from a plurality of sets of integrated values obtained by the pixel value integrating means, It is characterized in that the weight value of the aperture is obtained by a predetermined function or table.

According to a third aspect of the present invention, in the image processing apparatus according to the first aspect, the weight value setting means is a difference or difference between two of the integrated values obtained by the pixel value integrating means. It is characterized in that the weighted value of the aperture is obtained from the absolute value of A by a predetermined function or table.

According to a fourth aspect of the present invention, in the image processing apparatus according to the second or third aspect, the function or table of the weight value setting means is a distribution of integrated values obtained by scanning an image with the aperture. It is characterized in that it is set based on.

According to a fifth aspect of the invention, in the image processing apparatus according to any one of the second to fourth aspects, the function of the weight value setting means or the value in the table is the pixel value integrating means. It is characterized by having a correlation with the difference or ratio between the obtained integrated values.

According to a sixth aspect of the present invention, in the image processing apparatus according to the first to fifth aspects, the pixel value accumulating means is responsive to the integrated value of the pixel of interest and pixels in the vicinity thereof. It is characterized in that the size, shape, and number of openings are changed.

According to a seventh aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, the pixel value accumulating means has a plurality of apertures having different sizes, shapes and numerical apertures. A group is set, and the average value calculating means obtains a weighted average value for each aperture group, and obtains a restoration value based on them.

According to an eighth aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, a feature amount detecting means for detecting an image feature amount near the target pixel is provided. The pixel value integrating means changes the size, shape, and numerical aperture of the opening according to the feature amount detected by the feature amount detecting means.

According to a ninth aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, a feature amount detecting means for detecting an image feature amount near the target pixel is provided. The pixel value integrating means sets a plurality of aperture groups having different sizes, shapes, and numerical apertures, and the average value calculating means obtains a plurality of weighted average values for each aperture group,
The restoration value is obtained from the plurality of weighted average values based on the image feature amount detected by the feature amount detecting means.

According to a tenth aspect of the present invention, in the image processing apparatus according to the eighth or ninth aspect, the image feature amount is a filter processing result by a first-order differential filter.

[0017]

According to the present invention, in the input limited gradation data, each pixel is set as a pixel of interest in order, and the pixel value is integrated by the pixel value integrating means for each of the target pixel and a plurality of apertures set around it. To be done. The plurality of integrated values are sent to the weight value setting means, and the weight value of each aperture is calculated based on the values. Further, the plurality of integrated values are also sent to the average value calculating means, and the average value calculating means obtains a weighted average value from the integrated value and the weight value of each opening obtained by the weight value setting means, and this weighted average is calculated. The value is output as the restored value of the pixel of interest. As described above, according to the present invention, since the restoration is performed by the weighted averaging process for each aperture, a sufficient gradation can be obtained as compared with the conventional case where an image of a fixed size including a target pixel is used to smooth an image. The image can be reproduced with the number and the image quality.

According to the invention of claim 2, in the invention of claim 1, from the set of a plurality of integrated values,
The weighted value of each aperture is calculated according to the integrated pixel value by a predetermined function or table. For example, the weighted value of the opening not including the target pixel can be obtained from the set of the integrated value of the opening including the target pixel and the integrated value of the other openings. In such an example, the relationship between the integrated value of the opening including the target pixel and the integrated values of other openings, that is, the gradation change can be known in detail. With such a configuration, it is possible to perform gradation restoration processing corresponding to a complicated gradation change in the image.

According to a third aspect of the present invention, in the first aspect of the invention, each aperture is formed by a predetermined function or table based on the difference of two integrated values or the absolute value of the difference. Since the weighted value of is calculated, the tone restoration processing can be performed with a simpler function or a smaller number of tables.

These functions or tables are defined in claim 4.
As described above, since it can be set based on the distribution of the integrated value obtained by scanning the image with the aperture used for the restoration processing, the statistical between the multi-valued image and the pseudo halftone image is It is possible to perform gradation restoration processing that reflects the relationship.

Further, as in the invention described in claim 5,
By setting the function that calculates the weighted value or the value in the table so that it has a correlation with the difference or ratio between the integrated values, the weighted value is set according to the density difference between the openings or the density ratio. It is possible to obtain a good gradation restoration image in which blurring at the edge portion is suppressed.

Further, in the invention described in any one of claims 1 to 5, according to the integrated value in the vicinity of the pixel of interest, as in the invention described in claim 6, the size of the aperture group used for the restoration processing, By changing the shape and the numerical aperture, it is possible to suppress the residual of the pseudo-halftone pattern in the highlight portion of the image that is visually noticeable, and to perform a good gradation restoration process. Alternatively, as in the invention according to claim 7, by using a plurality of aperture groups having different sizes, shapes, and numerical apertures, and obtaining a restoration value from a plurality of weighted average values for each aperture group, similarly, It is possible to suppress the residual of the pseudo-halftone pattern in the highlight portion of the image that is visually noticeable and to perform a good gradation restoration process.

Further, in the invention according to any one of claims 1 to 5, the feature amount of the image in the vicinity of the target pixel, for example,
The size of the aperture group used in the restoration process is detected according to the detected feature amount by detecting the primary differential filter processing result and the like as in the invention described in claim 10. By changing the shape and the numerical aperture, it is possible to perform good gradation restoration processing according to the characteristics of a given image. Alternatively, as in the invention according to claim 9, by using a plurality of aperture groups having different sizes, shapes, and numerical apertures, and obtaining a restoration value from a plurality of weighted average values for each aperture group based on the feature amount Also, similarly, it is possible to perform good gradation restoration processing according to the characteristics of the given image.

[0024]

FIG. 1 is a block diagram showing a first embodiment of an image processing apparatus according to the present invention. In the figure, 1 is an image input unit, 2 is an image storage unit, 3 is an image output unit, 4 is a gradation restoration processing unit, 5 is a control unit, 11 is a pixel value integration unit, and 12 is a unit.
Is a weight value setting unit, and 13 is an average value calculation unit. The image input unit 1 inputs the limited gradation data transmitted or accumulated. The image storage unit 2 stores the image data input from the image input unit 1. The image output unit 3 is composed of an output device such as a multi-valued printer and outputs the processing result. The gradation restoration processing unit 4 performs gradation restoration processing. The control unit 5 controls each unit.

The tone restoration processing unit 4 has a pixel value integration unit 11, a weight value setting unit 12, and an average value calculation unit 13. The pixel value integration unit 11 integrates pixel values for each of a plurality of openings set around the pixel of interest. The weight value setting unit 12 sets a weight value for each integrated value based on the plurality of integrated values calculated by the pixel value integrating unit 11. The average value calculation unit 13 performs weighted averaging processing based on the plurality of integrated values calculated by the pixel value integration unit 11 and the weight value corresponding to each integrated value set by the weight value setting unit 12 Find the pixel reconstruction value.

In FIG. 1, image data is input from the image input unit 1, stored in the image storage unit 2, read out by the tone restoration processing unit 4 and subjected to tone restoration processing, and image output is performed in the order of processing. It is sent to the section 3 and output.

FIG. 2 is a flow chart showing an example of the processing flow of the gradation restoration processing unit 4 in the first embodiment of the image processing apparatus of the present invention. In FIG. 2, the width and height of the image to be processed are W pixels × H lines, the upper left of the image is (0, 0), and the lower right is (W-1, H-1). S
At 31, the target pixel is set to the upper left pixel of the input image, and the gradation restoration process is started.

In step S32, image data for p × q pixels centering on the pixel of interest is read from the image storage unit 2 into the pixel value integration unit 11. The read image data for p × q pixels needs to have a size enough to include all the openings set by the pixel value integration unit 11. The pixel value integration unit 11 sets a plurality of openings around the target pixel and integrates the pixel values included in each opening.

In S33, the plurality of integrated values obtained in S32 are transferred to the weight value setting unit 12, and the weight value setting unit 12
Calculate the weighted value for each aperture. The method of setting the weight value will be described later.

In step S34, the integrated value for each opening obtained in step S32 and the weighted value for each opening obtained in step S33 are transferred to the average value calculator 13, and a weighted averaging operation is performed from the integrated value and the weighted value. Then, the restoration value of the pixel of interest is calculated. Then, in S35, the restoration value of the pixel of interest obtained in S34 is sent to the image output unit 3 and sequentially output.

In S36, it is determined whether or not the process for one line of pixels is completed. If the line is being processed, the target pixel is changed to the pixel to the right of the current pixel in S37, Return to S32. If the processing of the line is completed, that is, if the current pixel of interest is the rightmost edge,
In S38, it is determined whether the processing has been performed up to the final line. If it is not the final line, the leftmost pixel one line below is taken as the pixel of interest in S39, and the process returns to S32. When the processing of the final line is completed, the processing of the gradation restoration processing unit 4 is completed.

An example of the operation of the first embodiment of the image processing apparatus of the invention will be described below based on a specific example.
In the following description, the case where binary data is restored to 256 values will be described as an example. However, even when data of three or more values is input as input, and when the gradation of output data is other than 256 values, It can be processed similarly.

FIG. 3 is an explanatory diagram of an example of an opening set in a specific example of the first embodiment of the present invention. In FIG. 3, the black dot in the center is the target pixel. Here, nine 3 × shown by hatching in FIG.
Three areas are set as openings. The pixel value integration unit 11 reads 9 × 9 pixels including all openings and integrates the pixel values in each set opening. When the input is a binary image, this integration is performed by counting the number of white or black pixels in the 3 × 3 area. Here, the central opening including the pixel of interest is block0, and the other openings are block.
1 to block8, and integrated values of black pixels in the block0 to block8 regions are referred to as p0 to p8.

Nine integrated values p obtained by the pixel value integrating section 11
0 to p8 are transferred to the weight value setting unit 12. In the weight value setting unit 12, two integrated values p0 and pk (k = 1 to 8)
For example, a table in which the values shown in the graph of FIG. 7 are stored is subtracted from the combination of 1 to obtain the weighted values of 0 to 1 corresponding to p1 to p8. Further, the weight value corresponding to p0 is set to 1 here. Hereinafter, these weighted values are represented by w0 to w8, respectively. The weight value used here is predetermined, and the method for determining it is as follows.
It will be described later together with the graph shown in FIG. 7.

Nine integrated values p obtained by the pixel value integrating unit 11
0 to p8 and the nine weight values w0 to w8 obtained by the weight value setting unit 12 are transferred to the average value calculation unit 13. The average value calculation unit 12 obtains a weighted average value by the following equation (1), and outputs the result to the image output unit 3 as the restored value of the pixel of interest. (Restored value) = (w0 × p0 + w1 × p1 + ... + w7 × p7 + w8 × p8) / (w0 + w1 + ... + w7 + w8) × (255/9) = (p0 + w1 × p1 + ... + w7 × p7 + w8 × p8) / (1 + w1 + ... + w7 + w8) (255/9)-(1)

When there is no difference in the integrated value of each aperture, it is considered that there is little change in density around it, so smooth multivalue restoration can be performed by performing smoothing as a whole. Also,
If there is a difference in the integrated value, it is considered that there is a density change between the openings or in the vicinity thereof. Therefore, the edges can be preserved by not using the openings having the differences for smoothing. However, when the smoothing method is switched in this way, unnaturalness occurs due to a sudden change in the smoothing characteristic at the switching part, so a weighting function is set to continuously change the smoothing characteristic, and edge preservation and smoothing are performed. Is smoothly switched.

In this example, the restoration is performed using the opening as shown in FIG. 3, but the position, size and shape of the opening are not limited to this. FIG. 4 is an explanatory diagram of another example of the opening set in the specific example of the first embodiment of the present invention. For example, in FIG. 4A, nine openings of 5 × 5 pixels are provided. Further, in FIG. 4B, 16 openings of 4 × 4 pixels are provided. In this example, the pixel of interest is not in the center of the region of 16 × 16 pixels including the entire opening. In FIG. 4C, eight 3 × 3 pixel areas are arranged around the modified 21 pixel area including the target pixel. Further, although these areas are adjacent to each other, areas apart from each other may be set. In this way, the way of taking the opening can be variously changed.

For example, when apertures having different aperture sizes are set as shown in FIG. 4 (C), weighted averaging is performed by an equation in which the size difference is corrected as in the following equation (2). Need to do. (Restored value) = {p0 × (9/21) + w1 × p1 + ... + w7 × p7 + w8 × p8} / (1 + w1 + ... + w7 + w8) × (255/9)-(2)

In the above description, the average value calculating section 13 calculates the weighted average value by using the equations (1) and (2), but it is necessary to use a modified configuration of these within a range that produces a similar result. You can also For example, in the above description, the average value calculation unit 13 performs normalization by the sum of weight values and multiplication for setting the output value range to 0 to 255. You can also do it at. Specifically, the weight value setting unit 12 calculates w0 to w8 and then converts and outputs each weight value by the formula (3), and the average value calculation unit 13 calculates the restored value by the formula (4). Even if it is configured, the same result as the expression (1) can be obtained. wi '= wi / (1 + w1 + ... + w7 + w8) * (255/9) (i = 0-8)-(3) (restoration value) = w0' * p0 + w1 '* p1 + ... + w7' * p7 + w8 '* p8- ( 4)

FIG. 5 is a block diagram showing an example of the average value calculation unit 13 in the first embodiment of the image processing apparatus of the present invention. In the figure, 21 is a total calculation unit, 22 is a table storage unit, 23 is a multiplication processing unit, and 24 is a multiplication value integration unit. The calculation in the average value calculation unit 13 such as the formula (1) and the formula (2) described above includes multiplication and division. Here, by multiplying each weight value by the integrated value using a table and switching the table for multiplication according to the total sum of weight values, it is possible to exclude division processing that has a large processing load. is there. FIG. 5 shows an example of the configuration when processing is performed using a multiplication table.

The number of black pixels in each opening accumulated by the pixel value accumulating unit 11 is output to the weight value setting unit 12 and the average value calculating unit 13 as 4-bit data of 0-9. The weight value setting unit 12 obtains nine weight values from the integrated value and outputs the nine weight values to the average value calculation unit 13. This weighted value is, for example, 0
It is assumed that a decimal value of ˜1 is quantized into an integer value of 0 to 255 and is output in the form of 8-bit data. Average value calculator 1
In 3, the internal sum total calculation unit 21 obtains the total sum of the quantized weight values wk (k = 0 to 8), and the table storage unit 2
Send to 2. The table storage unit 22 uses the total sum of the weight values input from the total sum calculation unit 21 to generate a table T (w, p), and the multiplication processing unit 23
The table T (w, p) is output to. (Output value) = T (w, p) = w × p / (w0 + w1 + ... + w7 + w8) × (255/9) − (5)

The multiplication processing section 23 receives this table, subtracts the table from the set of the integrated value and the weighted value, and outputs the result to the multiplied value integrating section 24. The multiplication value integration unit 24 integrates the outputs corresponding to the numerical aperture (here, 9) from the multiplication processing unit 23, and outputs the integration result as a restored value to the image output unit 3 or the like. With such a configuration, the gradation restoration process can be performed without using the division process.

Next, a method of determining the weight value by the weight value setting section 12 will be described. Number of black pixels p0 and p in two openings
When k is equal, it can be assumed that the two apertures belong to similar density regions on the multi-valued image that is the basis of the processed image. If p0 and pk are not equal, this 2
The probability that the two apertures belonged to similar density regions was p0.
It is considered that the relationship is inversely proportional to the absolute value of the difference between pk and pk. Considering that an amount corresponding to this possibility is set as the weight value wk for pk, it can be statistically obtained by the following method, for example.

A large number of sets of multi-valued images and images obtained by performing pseudo-halftone processing on the multi-valued images are prepared, and the pseudo-halftone images are scanned with an aperture having the same size as the previous aperture group, and the original multi-value image corresponding to each integrated value is obtained. The frequency distribution of the value pixel values is obtained, and this is normalized. The normalized frequency statistically represents what kind of pixel value the pseudo halftone area showing each integrated value had on the original image. When the area of overlap of the normalized frequency distributions is obtained, it becomes the weight value wk corresponding to the set (p0, pk) of integrated values.

FIG. 6 is an explanatory diagram of an example of setting the weight value in a specific example of the first embodiment of the present invention. Figure 6
Then, assume that the original multi-valued image has 256 gradations, and 2
The figure shows a case where the image subjected to the binarization processing is scanned by an opening having an opening size of 3 × 3. At this time, since the integrated value of the aperture is 0 to 9, the frequency of the pixel value of the multivalued image corresponding to the integrated value of the opening has a distribution as shown in FIG. 6A, for example. Here, the integrated value p0 is 3 and the integrated value pk is 5
The distribution in the case of is shown. It shows how much the overlapping portions of these two distributions may have the same density region. In FIG. 6A, the area of the cross-hatched portion or the ratio of the areas is such that the integrated value p0 is 3
Is the weighted value w5. This weighted value changes depending on the integrated value p0. A set of integrated value p0 and integrated value pk (p
F (p0, p), where F (p0, pk) is the area or ratio of the overlapping frequency distributions in (0, pk)
The value of k) can be calculated, for example, as shown in FIG.

FIG. 7 is an explanatory diagram of an example of the weight value function used in the weight value setting unit 12 in the specific example of the first embodiment of the present invention, and FIG. 8 is used when the processing target image is generated. It is explanatory drawing of an error diffusion coefficient. Weighting in the case where the binarization processing is performed using the error diffusion coefficient as shown in FIG. 8 on the basis of the 256-gradation multivalued image, and the obtained binarized image is the target image. The values are shown in FIG. In FIG. 7, by the method described in FIG. 6, F (p0,
pk), which is shown in the graph. In the figure, the thin solid line is for p0 = 0, the thin broken line is for p0 = 1, the thin dotted line is for p0 = 2, and the thin dash-dotted line is for p0.
= 3, a thin two-dot chain line is for p0 = 4, a thick solid line is for p0 = 5, a thick broken line is for p0 = 6, and a thick dotted line is for p0 = 7. The dashed line is p0
= 8, and the thick two-dot chain line shows the case of p0 = 9.

Here, the method of obtaining the weight value for a specific pseudo halftone method has been described. However, the weight value thus obtained is binarized using different error diffusion coefficients,
It also produces valid results for other pseudo-halftone schemes. Further, by obtaining weight values for a plurality of pseudo halftone methods and performing an operation such as averaging those weight values, a weight value table that is more effective for many methods can be obtained.

Thus obtained F (p0, p
The value of k) is, for example, as a table, the weight value setting unit 12
Can be kept at. Alternatively, a graph as shown in FIG. 7 is approximated by an appropriate function, and the weight value setting unit 12
A mechanism for generating a function value may be used.

FIG. 9 is an explanatory diagram of a modification of the setting of the weight value in the specific example of the first embodiment of the present invention. As can be seen from FIG. 7, the shape of each function does not change significantly with respect to the change of p0. Therefore, as shown in FIG. 9A, the functions existing corresponding to the respective values of p0 are superposed and approximated, and as shown in FIG. 9B, F (p0, p
By changing k) to F (p0-pk), the table capacity and the like can be reduced. Furthermore, this function F
When (p0-pk) is close to the left-right symmetry, FIG.
It is also possible to further reduce as F (| p0-pk |) as shown in FIG.

FIG. 10 is an explanatory diagram of another modification of the setting of the weight value in the specific example of the first embodiment of the present invention. It is known that human vision has a property that a pseudo halftone pattern remaining in a bright portion is more noticeable than a dark portion remaining. In consideration of such a phenomenon, as shown in FIG. 10B, the weight value table corresponding to the bright portion is intentionally widened so that the pseudo halftone pattern removal effect is enhanced in the bright portion. It is also possible to devise.

The above description has been made on the premise that a multi-valued image with 256 gradations is binarized using the error diffusion coefficient as an image to be processed. The weighted value can be obtained by the same method even for images other than the values, the output other than the 256 gradations, or the opening other than the 3 × 3 pixels as shown in FIG.

Next, a second embodiment will be described. In the above-described first embodiment, restoration is performed using the same aperture group for each pixel of interest. However, by changing the parameters such as aperture size and position according to the characteristics of each pixel of interest, adaptive processing is performed on the edges and flats, and more edges are saved to improve the pseudo halftone pattern. Can be removed. In the second embodiment, an example in which the aperture size is changed from the integrated pixel value in the vicinity of the pixel of interest will be described.

FIG. 11 is an explanatory diagram showing an example of gradation restoration processing in the second embodiment of the image processing system of the invention. As described above, human vision has the property that the pseudo halftone pattern remaining in the bright portion is more noticeable than the remaining in the dark portion. Therefore, for example, an opening including a target pixel as shown in FIG. 11 is set, pixel values are integrated, and the size of the opening for obtaining the weighted average value is switched according to the integrated value. In FIG. 11, as an opening including a target pixel, the number of black pixels therein is integrated by an opening of 5 × 5 pixels, and the integrated value p is equal to or larger than a certain threshold Th, that is, in a dark portion, as shown in FIG. 3, for example. Na 3x3
When the integrated value p is smaller than a certain threshold Th, an aperture having a pixel size of 4 × 4 pixels is used. As described above, by making the opening size in the bright portion larger than the opening size in the dark portion, the smoothing is strongly performed in the bright portion, and the residual pseudo halftone pattern can be reduced.

FIG. 12 is a flow chart showing an example of the processing flow of the gradation restoration processing section 4 in the second embodiment of the image processing apparatus of the invention. In FIG. 12, the width and height of the image are W pixels × H lines, the upper left of the image is (0, 0), and the lower right is (W-1, H-1). Also,
As a specific example, an opening as shown in FIG. 11 is set. In S41, the target pixel is set to the upper left pixel of the input image, and the gradation restoration process is started.

In step S42, image data of a predetermined pixel size centered on the pixel of interest, here 5 × 5 pixels, is read from the image storage unit 2 into the pixel value integration unit 11, and the number of black pixels contained therein is integrated. . This integrated value is called p.

In S43, the integrated value p obtained in S42
Is compared with a predetermined threshold Th. As a result, Th
If ≦ p, the process proceeds to S44, and if p <Th, the process proceeds to S46. When Th ≦ p, it indicates that a relatively large number of black pixels are present in the vicinity of the pixel of interest, that is, it is a dark region. Therefore, a gradation restoration process with a weak degree of smoothing is performed using an opening of 3 × 3 pixels. In S44,
Image data for 9 × 9 pixels including an opening of 3 × 3 pixels centered on the pixel of interest is read from the image storage unit 2 into the pixel value integration unit 11, and the pixel value is calculated for each of the nine openings of 3 × 3 pixels. Is added. In S45, the nine integrated values obtained in S44 are transferred to the weight value setting unit 12, and the weight value setting unit 12 obtains the weight value for each opening.

When p <Th, it means that there are not many black pixels in the vicinity of the pixel of interest, that is, the area is bright. Therefore, gradation restoration processing with a high degree of smoothing is performed using a 4 × 4 aperture. In step S46, image data for 12 × 12 pixels including all pixels of nine 4 × 4 pixel openings is read from the image storage section 2 into the pixel value integration section 11, and pixel values are calculated for each 4 × 4 pixel opening. Add up. S
In 47, the nine integrated values obtained in S46 are transferred to the weight value setting unit 12, and the weight value setting unit 12 obtains the weight value for each aperture.

The nine integrated values and the weighted values thus obtained are sent to the average value calculation unit 13, the restored value of the pixel of interest is calculated by the weighted average processing in S48, and S49.
At, the restored value is transferred to the image output unit 3 and sequentially output.

Then, in S50, it is determined whether or not the processing for the pixels of one line is completed. If the processing for the line is in progress, the target pixel is changed to the pixel to the right of the current pixel in S51. Then, the process returns to S42. If the processing of the line is completed, that is, if the current pixel of interest is the rightmost end, it is determined in S52 whether processing has been performed up to the final line. If it is not the last line, S53
, The leftmost pixel one line below is taken as the pixel of interest, and S
Return to 42. When the processing of the final line is completed, the processing of the gradation restoration processing unit 4 is completed.

Since the larger the opening size at the time of restoration, the stronger the smoothness is, the processing like this causes the target pixel having a small number of black pixels (= belonging to a bright area) to be a smooth halftone pattern with a strong smoothing. The target pixel, which is well removed and has a large number of black pixels in the periphery (= belongs to a dark area), is better smoothed and the edges are better preserved.

Although the aperture size is changed from the integrated pixel value in the vicinity of the pixel of interest here, the present invention is not limited to this, and any combination of aperture groups having different smoothing characteristics can be used. It can be something like this. For example, changing the number of apertures set in the vicinity of the pixel of interest without changing the aperture size, or changing both the aperture size and the number of apertures,
Alternatively, various methods such as changing the shape of the opening or changing the position of the opening can be considered.

FIG. 13 is an explanatory diagram of another example of the gradation restoration processing in the second embodiment of the image processing apparatus of the invention. In the above example, the degree of smoothing does not differ so much, so the openings are switched by thresholding. However, when such switching is performed, the change in image quality may be noticeable at the switching portion. In such a case, as shown in FIG. 13, it is possible to make the change in image quality inconspicuous by performing calculation based on a plurality of restoration results using a plurality of types of openings. In the example shown in FIG. 13, when the integrated value p in FIG. 11 is equal to or larger than the threshold Th, the restored value 1 obtained by performing the restoration process using nine openings of 3 × 3 pixels, and the integrated value p are smaller than the threshold Th. The ratio of contribution is changed according to the integrated value p by using the restored value 2 obtained by performing the restoration process using the nine apertures of 4 × 4 pixels. For example, as shown in FIG. 13, the restoration value can be calculated by (restoration value) = {(restoration value 1) × p + (restoration value 2) × (25−p)} / 25.

Next, a third embodiment will be described. In the above-described second embodiment, the size of the opening used for restoration is changed according to the integrated pixel value in the vicinity of the pixel of interest. Here, as another specific example, an example in which the opening shape is changed according to the feature amount near the target pixel will be described.

FIG. 14 is a block diagram showing a third embodiment of the image processing apparatus of the invention. In the figure, the same parts as those in FIG. 6 is a feature amount detection unit. The feature amount detection unit 6 detects the feature amount of the image near the pixel of interest. Various features can be considered. Here, as an example of the feature amount, it is assumed that the edge feature in the image is detected in order to adaptively process the edge portion and the flat portion. For the detection of the edge portion, for example, a first-order differential filter can be used. The characteristic amount detected by the characteristic amount detecting unit 6 is transmitted to the gradation restoration processing unit 4.

FIG. 15 is an explanatory diagram of an example of the first-order differential filter. Here, a filter having a size of 7 × 7 pixels is shown as an example. Using the value of this filter, 7 ×
The sum of products with the pixel value of 7 pixels can be calculated, and the value can be used as the feature amount. In the first-order differential filter shown in FIG. 15 (A), when the left side and the right side of the pixel of interest differ greatly in density, the value of the product-sum calculation becomes large. Therefore, it is possible to detect a vertically existing density boundary. Similarly,
The first-order differential filter shown in FIG. 15 (B) can detect the density boundary that exists laterally. The first-order differential filter is not limited to the size of 7 × 7, but may be any size as long as the feature can be captured.

FIG. 16 is an explanatory diagram of an example of a plurality of aperture groups set in the third embodiment of the invention. For example, when it is detected that a density boundary exists in the vertical direction by a first-order differential filter as shown in FIG. 15,
The horizontal smoothing process is weakened to preserve the vertical density boundary. Therefore, for example, an opening of 3 × 5 pixels as shown in FIG. 16A is set. Further, when it is detected that the density boundary exists in the horizontal direction, the smoothing process in the vertical direction is weakened to save the density boundary in the horizontal direction. Therefore, for example, an opening of 5 × 3 pixels as shown in FIG. 16B is set. Note that in this embodiment, when the density boundary in either direction is not detected, an opening of 4 × 4 pixels is used as shown in FIG. 16C.

FIG. 17 and FIG. 18 are flowcharts showing an example of the flow of gradation restoration processing in the third embodiment of the image processing apparatus of the invention. Also in FIG. 17, the width and height of the image are W pixels × H lines, the upper left of the image is (0, 0), and the lower right is (W-1, H-1). Further, the 7 × 7 pixel filter shown in FIG. 15 is used to extract the feature amount, and the aperture set by the pixel value integration unit 11 is the one shown in FIG. 16 as a specific example. In S61, the target pixel is set to the upper left pixel of the input image, and the gradation restoration process is started.

In S62, image data for 7 × 7 pixels centering on the pixel of interest is read from the image storage unit 2 into the feature amount detection unit 6. Then, in the feature amount detection unit 6, FIG.
Differentiation is performed using a first-order differential filter as shown in (A) and (B), and the absolute value of each result is output. As described above, the density boundary in the vertical or horizontal direction can be detected by using the first-order differential filter as shown in FIG. Since the sign of the primary differential value differs depending on the direction of density change, the absolute value of these primary differential values is calculated and always output as a positive value. Here, FIG.
The absolute values of the first-order differential values obtained by the filters shown in (a) and (b) are qa and qb, respectively.

In S63, qa, q obtained in S62
b is compared with a predetermined threshold Th. As a result, q
If both a and qb are less than or equal to the threshold value, then in step S64,
If either b is larger than the threshold value, the process proceeds to S66.

When both the values of qa and qb are less than or equal to the threshold value Th, it is considered that there is no large density change in the vertical and horizontal directions around the pixel of interest. So, for example,
Using 9 apertures of 4 × 4 pixels shown in FIG.
Performs an isotropic restoration process vertically and horizontally. In step S64, image data for 12 × 12 pixels including all the pixels of nine 4 × 4 pixel openings shown in FIG. 16C is read from the image storage section 2 into the pixel value integration section 11, and for each opening. The pixel value is added to. Then, in S65, the nine integrated values obtained in S64 are transferred to the weight value setting unit 12, and the weight value setting unit 1
In step 2, the weight value for each aperture is calculated.

Either the value of qa or qb is the threshold Th.
If it is larger than that, it is considered that there is a large density change in the vertical or horizontal direction around the pixel of interest. Therefore,
For example, the restoration process according to the direction of density change is performed using the 3 × 5 pixel aperture shown in FIG. 16A or the 5 × 3 pixel aperture shown in FIG. 16B. In S66, in order to see whether the density changes in the vertical direction or the horizontal direction, qa
And qb are compared. If qa ≧ qb, S67
If qa <qb, the process proceeds to S69. When the process proceeds to S67, image data for 9 × 15 pixels centered on the pixel of interest is read from the image storage unit 2 into the pixel value integration unit 11, and nine 3 × 5 pixels shown in FIG. Pixel values are integrated for each aperture. Then, in S68, the nine integrated values obtained in S67 are transferred to the weight value setting unit 12, and the weight value setting unit 12 obtains the weight value for each opening.

When the process proceeds to S69, the image data of 15 × 9 pixels centered on the pixel of interest is read from the image storage unit 2 into the pixel value integration unit 11, and the nine image data 5 shown in FIG. The pixel value is integrated for each opening of × 3 pixels. And
In S70, the nine integrated values obtained in S69 are transferred to the weight value setting unit 12, and the weight value setting unit 12 obtains the weight value for each opening.

The nine integrated values and weighted values thus obtained are sent to the average value calculation unit 13, and the weighted average processing calculates the restored value of the pixel of interest in S71, and S72
At, the restored value is transferred to the image output unit 3 and sequentially output.

After that, in S73, it is determined whether or not the processing for the pixels of one line is completed. If the processing for the line is in progress, the pixel of interest is changed to the pixel to the right of the current pixel in S74. Then, the process returns to S62. If the processing of the line is completed, that is, if the current pixel of interest is the rightmost end, it is determined in S75 whether the processing has been performed up to the final line. If it is not the last line, S76
, The leftmost pixel one line below is taken as the pixel of interest, and S
Return to 62. When the processing of the final line is completed, the gradation restoration processing is completed.

As described above, according to the change in density around the pixel of interest, the shape of the opening is narrowed in the direction where the density change is strong, and the opening is widened in the direction where the density change is weak. Images can be obtained. Although the first-order differential filter that detects the density boundaries in the vertical and horizontal directions is used here, it is also possible to add a filter corresponding to the diagonal direction or a filter having a different size to detect the finer edges to change the aperture. It is possible. In addition, it is expected that a good restored image can be obtained by using the output of a secondary differential filter other than the primary differential filter, or the feature amount such as the number of pixel value inversions when the input image is binary. it can.
In addition, the size of the opening set according to the feature amount may be changed, and the position, shape, and number may be changed.

As described with reference to FIG. 13 in the second embodiment, a plurality of aperture groups having different sizes, positions, shapes, and numbers of apertures are set, and the restoration value is calculated for each aperture group. After that, it is possible to increase / decrease the ratio of the restoration value obtained from each aperture group based on the detection result of the feature amount detection unit 6 to obtain the final restoration value.

[0077]

As is apparent from the above description, according to the present invention, the weighted average processing according to the integrated value of the pixel values by a plurality of apertures makes it possible to compare with the conventional simple averaging by one aperture. It is possible to restore a good gradation while suppressing blurring of the edge portion and residual pseudo halftone pattern. Further, since the weighted averaging process is used, it is possible to realize an image processing device capable of reproducing sufficient gradations as compared with the restoration by the simple averaging process. Furthermore, by changing the aperture according to the characteristics of the image near the pixel of interest, or by using multiple restoration values by each aperture group, it is possible to suppress the residual pattern of the highlight portion and blurring of the edge portion. Therefore, there is an effect that it is possible to realize an image processing apparatus capable of obtaining a better gradation restored image.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating a first embodiment of an image processing apparatus according to the present invention.

FIG. 2 is a flowchart showing an example of a processing flow of a gradation restoration processing unit 4 in the first embodiment of the image processing apparatus of the present invention.

FIG. 3 is an explanatory diagram of an example of an opening set in a specific example according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram of another example of the opening set in a specific example according to the first embodiment of the present invention.

FIG. 5 is a configuration diagram showing an example of an average value calculation unit 13 in the first embodiment of the image processing apparatus of the present invention.

FIG. 6 is an explanatory diagram showing an example of setting a weight value in a specific example of the first exemplary embodiment of the present invention.

FIG. 7 is an explanatory diagram of an example of a weight value function used in the weight value setting unit 12 in the specific example of the first exemplary embodiment of the present invention.

FIG. 8 is an explanatory diagram of an error diffusion coefficient used when generating an image to be processed.

FIG. 9 is an explanatory diagram of a modified example of setting a weight value in a specific example of the first embodiment of the present invention.

FIG. 10 is an explanatory diagram of another modification of setting the weight value in the specific example of the first embodiment of the present invention.

FIG. 11 is an explanatory diagram of an example of gradation restoration processing in the second embodiment of the image processing apparatus of the invention.

FIG. 12 is a flowchart showing an example of the processing flow of the gradation restoration processing unit 4 in the second embodiment of the image processing apparatus of the present invention.

FIG. 13 is an explanatory diagram of another example of gradation restoration processing in the second embodiment of the image processing apparatus of the invention.

FIG. 14 is a block diagram showing a third embodiment of the image processing apparatus of the invention.

FIG. 15 is an explanatory diagram of an example of a first-order differential filter.

FIG. 16 is an explanatory diagram of an example of a plurality of aperture groups set in the third embodiment of the invention.

FIG. 17 is a flowchart showing an example of the flow of gradation restoration processing in the third embodiment of the image processing apparatus of the invention.

FIG. 18 is a flowchart (continuation) showing an example of the flow of gradation restoration processing in the third embodiment of the image processing apparatus of the invention;

[Explanation of symbols]

1 is an image input unit, 2 is an image storage unit, 3 is an image output unit, 4
Is a gradation restoration processing unit, 5 is a control unit, 6 is a feature amount detection unit, 1
1 is a pixel value integration unit, 12 is a weight value setting unit, 13 is an average value calculation unit, 21 is a sum calculation unit, 22 ... Table storage unit, 2
3 ... Multiplication processing unit, 24 ... Multiplication value integrating unit.

Claims (10)

[Claims] 1. A pixel value integrating means for setting a plurality of openings around a target pixel and its surroundings and integrating pixel values for each opening, and a weight for each opening from the integrated value obtained by the pixel value integrating means. A weighted value setting means for obtaining a value, an average value calculation means for obtaining a weighted average value from the integrated value obtained by the pixel value integrating means and the weighted value obtained by the weighted value setting means are provided. An image processing apparatus, wherein each pixel is used as the target pixel, and gradation restoration processing is performed by the pixel value integration means, the weight value setting means, and the average value calculation means. 2. The weight value setting means obtains the weight value of the aperture by a predetermined function or table from a plurality of sets of the integrated values obtained by the pixel value integrating means. The image processing device according to claim 1. 3. The weight value setting means calculates the weight value of the aperture by a predetermined function or table from the difference or the absolute value of the difference between the two integrated values obtained by the pixel value integrating means. The image processing apparatus according to claim 1, which is obtained. 4. The function or table of the weight value setting means is set based on a distribution of integrated values obtained by scanning an image with the aperture.
An image processing apparatus according to claim 1. 5. The function or the value in the table of the weight value setting means has a correlation with a difference or a ratio between the integrated values obtained by the pixel value integrating means. 5. The image processing device according to any one of items 4 to 4. 6. The pixel value integrating means changes the size, shape, and numerical aperture of the opening according to the integrated value of the pixel of interest and pixels in the vicinity thereof.
The image processing apparatus according to claim 5. 7. The pixel value accumulating means includes a size, a shape,
6. A plurality of aperture groups having different numerical apertures are set, and the average value calculating means obtains a weighted average value for each aperture group, and obtains a restoration value based on them. The image processing device according to any one of 1. 8. A feature quantity detecting means for detecting an image feature quantity in the vicinity of the pixel of interest is provided, and the pixel value integrating means has a size of the opening according to the feature quantity detected by the feature quantity detecting means. The image processing apparatus according to any one of claims 1 to 5, wherein the shape, the numerical aperture, and the numerical aperture are changed. 9. A feature quantity detecting means for detecting an image feature quantity in the vicinity of the target pixel, wherein the pixel value integrating means sets a plurality of aperture groups having different sizes, shapes, and numerical apertures, and the average value is set. The calculating means obtains a plurality of weighted average values for each aperture group, and obtains a restoration value from the plurality of weighted average values based on the image feature amount detected by the feature amount detecting means. The image processing apparatus according to any one of claims 1 to 5. 10. The image processing apparatus according to claim 8, wherein the image feature amount is a filter processing result by a first-order differential filter.