JP7344498B2 - Image processing device, image processing method and program - Google Patents

Description

The present invention relates to an image processing device, an image processing method, and a program for converting monochrome or color multi-value input image data into multi-value quantized image data.

The error diffusion method is widely used as a binary quantization method for printed images. The error diffusion method propagates the quantization error during binary quantization to the surrounding area, and reproduces a pseudo intermediate density by black and white density. Although the error diffusion method is easy to configure and control, it has the problem that the spatial variation in binary quantization of the contour portion of a printed image becomes large, and the contour portion cannot be accurately reproduced.

To solve this problem, the image processing device, program, and method described in Patent Document 1 solve the problem of reducing the outline of printed characters that occurs when using the error diffusion method, which is generally used as a quantization method for printed images. Quantization that preserves the smoothness of edge parts by quantizing edge parts and quantizing non-edge parts separately and then logically synthesizing them in order to reduce spatial variations in quantization. has been realized.

However, as shown in the embodiments of Patent Document 1 in FIGS. 10 and 11, Patent Document 1 targets characters and the like that have a clear brightness difference from the background, and for natural images such as photographs, Embodiment As shown in FIG. 5, only special areas with clear edges are processed, and the process is basically designed so as not to affect photographs. Therefore, it cannot be applied to preserving texture within an image, which is the object of the present invention.

In addition, although commercially available printers use binary quantized image data, in specialized printing fields such as newspapers and advertisements, binary quantized image data is not necessarily output, but rather is used to improve the expressiveness of gradation. In many cases, multi-level quantized images such as four-level images are printed.

Japanese Patent Application Publication No. 2018-98552

The present invention provides an image processing device, an image processing method, and a program that realize a method of converting monochrome or color multi-value input image data into multi-value quantized image data, and can be applied to emphasizing texture in images such as photographs. The purpose is to provide.

In order to solve the above problems, an image processing device of the present invention is an image processing device that converts multi-value input image data into multi-value quantized output image data, and detects texture information of the multi-value input image data. a texture detection unit; and an error diffusion unit that converts the multi-value input image data into the multi-value quantized output image data by performing tone conversion processing on the multi-value input image data so as to emphasize the texture information. An image processing device having a section.

It is preferable that the texture information is surface unevenness information, surface pattern information, brightness information, edge information, or contour information.

It is preferable that the image forming apparatus further includes an error distribution unit that calculates distribution of quantization errors of the conversion according to the texture information.

Preferably, the texture information is texture enhancement processing information of the multivalued input image data.

Preferably, the texture enhancement processing information is frequency subband information.

a difference calculation unit that generates a difference image between the multi-value input image data and the multi-value quantized image data quantized by the error diffusion unit; a visual processing unit that is a visual characteristic filter that inputs the difference image; further comprising a modification amount calculation unit that calculates a modification amount so as to preserve texture information, and a multivalued image modification unit that modifies the multivalued input image data based on the modification amount, and the multivalued image modification Preferably, the error diffusion section generates multilevel quantized output image data based on the output of the section.

The image processing method of the present invention is an image processing method for converting multi-value input image data into multi-value quantized output image data, and when the multi-value input image data has not been subjected to prior texture information processing, a texture detection step of detecting texture information of the multi-value input image data with a texture detection unit; and, if the multi-value input image data has undergone prior texture information processing, the texture information of the multi-value input image data is converting the multi-value input image data into the multi-value quantized output image data in an error diffusion unit by performing tone conversion processing on the multi-value input image data so as to emphasize the texture information using the multi-value input image data. An image processing method having the following steps.

The image processing method includes a step of generating a difference image between the multi-value input image data and the multi-value quantized image data quantized by the error diffusion section in a difference calculation section, and using a visual characteristic filter to generate the difference image. a step of inputting the multi-value input image data to a certain visual processing section, a step of calculating a correction amount in a correction amount calculation section so as to save the texture information, and a step of inputting the multi-value input image data in a multi-value image correction section based on the correction amount. Preferably, the method further includes the steps of correcting the image data, and generating the multi-value quantized output image data in the error diffusion section based on the output of the multi-value image correction section.

It is preferable that the process of calculating the correction amount in the correction amount calculation section is repeated repeatedly.

The program of the present invention is a program for causing a computer to function as each means of the image processing apparatus.

According to the present invention, an image processing apparatus, an image processing method, and an image processing apparatus that realize a method of converting monochrome or color multi-value input image data into multi-value quantized image data and can be applied to emphasizing texture in an image such as a photograph. This has the significant effect of being able to provide programs. Further, according to the present invention, by applying the process of the present invention to images such as CMYK, which are obtained by separating color images, processing of color multivalued images can also be realized.

1 is a block diagram showing a schematic configuration of one embodiment of an image processing device of the present invention. 1 is a flowchart showing the flow of an image processing method according to one embodiment of the present invention. FIG. 3 is a block diagram showing a schematic configuration of another embodiment of the image processing device of the present invention. 7 is a flowchart showing the flow of an image processing method according to another embodiment of the present invention. 1 is a schematic configuration diagram showing a printing device using an image processing device according to the present invention. These are natural images showing the results of Example 1 and Comparative Example 1, where (a) is the result of Comparative Example 1, and (b) is the result of Example 1.

Embodiments of the present invention will be described below, but these embodiments are shown by way of example, and it goes without saying that various modifications can be made without departing from the technical idea of the present invention. Note that the same members are represented by the same symbols.

In the figure, reference numerals 10A and 10B indicate respective embodiments of an image processing device according to the present invention. The image processing devices 10A and 10B are suitably used, for example, in a one-pass type inkjet printer 50 as shown in FIG. In FIG. 5, the inkjet printer 50 prints each color, for example, C (cyan), M (magenta), Y (yellow), and K (black), on the web-like printing base material 14 stretched between rolls 12a and 12b. Inkjet heads 16a, 16b, 16c, and 16d are provided for printing ink, and ink is ejected from the inkjet heads 16a, 16b, 16c, and 16d under the control of a head control unit 18.

The image processing apparatuses 10A and 10B are configured to output image data to the head control unit 18 so that printing is performed on the web-like printing base material 14 using inkjet heads 16a, 16b, 16c, and 16d.

<Image processing device 10A>
The schematic configuration of the image processing device 10A is shown in the block diagram of FIG. As shown in FIG. 1, the image processing device 10A is an image processing device that converts multi-value input image data 20 into multi-value quantized output image data 28, and detects texture information of the multi-value input image data 20. The texture detection unit 22 converts the multi-value input image data 20 into the multi-value quantized output image data by executing tone conversion processing on the multi-value input image data 20 so as to emphasize the texture information. The image processing device includes an error diffusion unit 26 that performs conversion.

The image processing device 10A is a computer including a control unit such as a CPU and a storage unit, and is caused to function as each means by a program to perform each process of the image processing device 10A. Moreover, each process of the image processing apparatus 10A may be performed via a network such as the Internet.

The multi-valued input image data 20 is a color or monochrome digital image expressed in 8-bit or other gradations, and is replaced with all non-binary images (for example, images expressed in 8-bit format). The multivalued quantized output image data 28 is quantized as follows.

One embodiment of an image processing method by the image processing device 10A will be described based on FIGS. 1 and 2. The image processing method can be executed by a program stored in a storage unit of a computer.

The image processing method of the present invention by the image processing device 10A is an image processing method for converting monochrome or color multi-value input image data 20 into monochrome or color multi-value quantized output image data 28. First, monochrome or color multi-value input image data 20 is input from another information terminal, storage medium, etc. (multi-value image data input step, S100).

It is determined whether the multivalued input image data 20 has been previously subjected to image processing that handles some texture information (for example, image processing that emphasizes texture) (preliminary texture information processing detection step, S102), and If the value input image data 20 has not undergone any texture information processing (for example, image processing that emphasizes texture) in advance, the texture information of the multi-value input image data 20 is detected by the texture detection unit 22. (texture detection step, S104).

Here, texture information means information on a texture part. Texture areas are physically areas where the brightness gradient in the image is high and the local area has a high brightness gradient dispersion, and the image features include fine surface irregularities and surface patterns with brightness changes, and fine surface patterns in a collection of fine objects. Refers to parts such as contours (edges). The texture information includes, for example, surface unevenness information, surface pattern information, brightness information, and contour information or edge information of a collection of fine objects of the texture portion. Note that these pieces of information are perceptually referred to as "unevenness," "pattern," "edge," etc., depending on the difference in their values regarding the luminance gradient and local luminance gradient dispersion information after noise removal.

In the texture detection step S104, for example, when the texture information is edge information, noise components are separated from the multi-value input image data by noise removal in advance, and then a luminance signal (brightness) is extracted from the multi-value input image data. The texture information is detected by extracting the edge strength using the gradient and calculating the texture strength using the local region variance of the extracted edge strength. Note that if the noise is not removed, the noise component will also be considered as a texture part, so it is preferable to separate it in advance.

In addition, if the multi-value input image data 20 has been previously subjected to image processing that handles some texture information (for example, image processing that emphasizes texture), the texture information of the multi-value input image data 20 (texture information utilization step, S106). In this case, since existing texture information is used, there is an advantage that the amount of calculation for detecting texture information is reduced.

The error distribution unit 24 also has a function of determining the distribution of quantization errors of the error diffusion method according to the strength of the texture information, and distributes the error so as to preferentially preserve the texture information. (error distribution step, S108).

Then, by executing tone conversion processing on the multi-value input image data so as to emphasize the texture information, the multi-value input image data is converted into the multi-value quantized output image data by the error diffusion unit 26. (Quantization step in the error diffusion section, S110). The error diffusion section 26 performs quantization based on the error distribution result determined by the error distribution section 24. Then, all pixels are repeatedly processed one by one by raster scanning, vector scanning, etc. (step of determining whether all pixels have been quantized or not, S112).

Once all pixels have been quantized, they are output as multi-value quantized output image data 28 (multi-value quantized output image data output step, S114). If there are remaining pixels, the process returns to the error distribution step S108 in the error distribution unit 24 and is repeated.

In addition, in the example of the image processing method by the image processing apparatus 10A described above, an example was shown in which the preliminary texture information processing detection step S102 was provided, but as another example of the image processing method by the image processing apparatus 10A, the preliminary texture information It is also possible to omit the process detection step S102 and perform the respective processes in cases where the multi-valued input image data 20 has been subjected to prior texture information processing and when prior texture information processing has not been performed.

<Image processing device 10B>
A schematic configuration of the image processing device 10B is shown in the block diagram of FIG. As shown in FIG. 3, the image processing device 10B is an image processing device that converts multi-value input image data 30 into multi-value quantized output image data 44, and detects texture information of the multi-value input image data 30. The texture detection unit 32 converts the multi-value input image data 30 into the multi-value quantized output image data 44 by performing tone conversion processing on the multi-value input image data 30 so as to emphasize the texture information. This is an image processing device having an error diffusion unit 42 that converts the image into an image.

The image processing device 10B further includes a difference calculation unit 36 that generates a difference image between the multi-value input image data 30 and the multi-value quantized image data quantized by the error diffusion unit 42, and a difference calculation unit 36 that receives the difference image. a visual processing section 38 that is a visual characteristic filter; a modification amount calculation section 34 that calculates a modification amount so as to preserve the texture information; and a multivalued image that modifies the multivalued input image data 30 based on the modification amount. The error diffusion unit 42 generates multi-value quantized output image data 44 based on the output of the multi-value image correction unit 40.

The image processing device 10B, like the image processing device 10A described above, is a computer equipped with a control unit such as a CPU and a storage unit, and is caused to function as each means by a program to perform each process of the image processing device 10B. . Furthermore, each process of the image processing device 10B may be performed via a network such as the Internet.

The multi-value input image data 30 is the same as the multi-value input image data 20, and is a digital image expressed in 8-bit or other gradation. The multivalued quantized output image data 44 is quantized so as to replace it with the expressed image).

One embodiment of an image processing method by the image processing device 10B will be described based on FIGS. 3 and 4. The image processing method can be executed by a program stored in a storage section of a computer, similarly to the image processing apparatus 10A described above.

The image processing method of the present invention by the image processing device 10B is an image processing method for converting monochrome or color multi-value input image data 30 into monochrome or color multi-value quantized output image data 44. First, monochrome or color multi-value input image data 30 is input from another information terminal, storage medium, etc. (multi-value image data input step, S200).

It is determined whether the multivalued input image data 30 has been previously subjected to image processing that handles some kind of texture information (for example, image processing that emphasizes texture) (preliminary texture information processing detection step, S202), and the multivalued input image data 30 is If the value input image data 30 has not been subjected to any image processing that handles texture information in advance (for example, image processing that emphasizes texture), the texture information of the multi-value input image data 30 is used as a texture. Detection is performed by the detection unit 32 (texture detection step, S204). The texture information and the texture detection step S204 are as described for the image processing apparatus 10A above.

Furthermore, if the multi-value input image data 30 has been subjected to image processing that handles some texture information in advance (for example, image processing that emphasizes texture), the texture information of the multi-value input image data 30 Use.

Further, the multi-value input image data 30, which has been subjected to image processing that handles some texture information in advance, is corrected by the multi-value image correction section 40 (multi-value image correction step, S212), and error diffusion is performed. Quantization is performed in the unit 42 (quantization step, S214).

The multi-value quantized image data quantized by the error diffusion unit 42 in S214 is sent to the difference calculation unit 36, where it is combined with the multi-value input image data 30 and the multi-value quantized image data quantized in S214. A difference image is generated by the difference calculation unit 36 (difference image generation step, S206).

This generated difference image is passed through a visual characteristic filter in the visual processing unit 38 (filter processing step, S208), and then in the correction amount calculation unit 34, the texture obtained by the texture detection unit 32 in the texture detection step S204 is The modification amount calculation unit 34 derives the modification amount so that the multi-level quantized image preserves the texture information (modification amount derivation step, S210).

Then, based on the amount of correction in the correction amount deriving step S210 by the correction amount calculating section 34, the input multi-value input image is corrected in the multi-value image correction section 40 (multi-value image correction step, S212), and this is Quantization is performed by the error diffusion unit 42 (quantization step, S214). Then, all pixels are repeatedly processed one by one by raster scanning or vector scanning (step of determining whether all pixels have been quantized, S216).

Once all pixels have been quantized, they are output as multi-value quantized output image data 44 (multi-value quantized output image data output step, S218). If there are remaining pixels, the process returns to the difference image generation step S206 in the difference calculation unit 36 and is repeated.

Note that in the example of the image processing method by the image processing device 10B described above, an example was shown in which the preliminary texture information processing detection step S202 was provided, but as another example of the image processing method by the image processing device 10B, the prior texture information processing detection step S202 is provided. It is also possible to omit the process detection step S202 and perform the respective processes when the multivalued input image data 30 has been subjected to prior texture information processing and when it has not been subjected to prior texture information processing.

By iteratively repeating the above process, quantization that perceptually preserves texture information can be achieved. The image processing method of the present invention using the image processing device 10B has the advantage that generally better quantization results can be obtained compared to the image processing method using the image processing device 10A.

The above-mentioned error diffusion units 26 and 42 perform quantization of all pixels in order by raster scanning, vector scanning, etc., pixel by pixel, using an error diffusion method. The error diffusion method is a type of method for smoothly expressing images by processing intermediate gradations. This is a method that minimizes the overall error by allocating errors that occur in the processing of pixels in a digital image to surrounding pixels, and then performing subsequent processing while taking into account the effects of allocating the error. As an example, a case will be described in which a four-value quantized image using C (cyan), M (magenta), Y (yellow), and K (black) inks is output.

For example, when C, M, Y, and K inks are used, each of CMYK has four values, and the four values are determined for each ink using the following method.
<In the case of C (cyan)>
output = 0 (dark cyan), if X ≦ a ₁ ,
output = 85 (medium cyan), if a ₁ < X ≦ a ₂ ,
output = 170 (light cyan), if a ₂ < X ≦a ₃ ,
output = 255 (no ink), if a ₃ < X
Here, a ₁ , a ₂ , and a ₃ are threshold parameters determined by the droplet size and ink concentration, and are determined by current color proofing techniques to minimize the color difference between the input and printed images. The four values for each of the other M, Y, and K inks are determined in the same manner as described above.

As long as it is smaller than the number of input gradations, the present invention is not limited to four-value quantization, but can also be applied to quantization of any number of gradations. Note that various factors in the printing process such as ink characteristics, paper characteristics, head characteristics, etc. accumulate and produce a non-linear effect. The four values will be determined by estimating the The error diffusion units 26 and 42 perform quantization of all pixels in order by raster scanning, vector scanning, etc., pixel by pixel, by diffusing errors based on this threshold parameter. Quantization is performed similarly for each of M, Y, and K inks other than C.

As another example, a case will be described in which a hexagonal quantized image using C (cyan), M (magenta), Y (yellow), and K (black) inks is output. For example, when C, M, Y, and K inks are used, each of CMYK has six values, and the six values are determined for each ink using the following method.
<In the case of C (cyan)>
output = 0 (dark cyan), if X ≦ a ₁ ,
output = 51 (slightly dark cyan), if a ₁ < X ≦ a ₂ ,
output = 102 (medium cyan), if a ₂ < X ≦a ₃ ,
output = 153 (slightly light cyan), if a ₃ < X ≦a ₄ ,
output = 204 (light cyan), if a ₄ < X ≦a ₅ ,
output = 255 (no ink), if a ₅ < X
Here, a ₁ , a ₂ , a ₃ , a ₄ , and a ₅ are threshold parameters determined by the droplet size and ink concentration, as described above. The error diffusion units 26 and 42 perform quantization of all pixels in order by raster scanning, vector scanning, etc., pixel by pixel, by diffusing errors based on these threshold parameters. The six values of M, Y, and K inks other than C are also determined in the above manner, and quantization is performed in the same manner.

FIG. 6 shows the printing results of photographic images according to Example 1 of the present invention and Comparative Example 1. In Example 1 and Comparative Example 1, printing was performed using an inkjet printer FXIJ manufactured by Think Laboratory Co., Ltd. using a 600 dpi inkjet KJ48 for water-based inks, and using C, M, Y, and K water-based inks. . The droplet sizes used were 0 pL (ink ejection amount: zero picoliters, no dots), 3 pL, 4 pL, and 5 pL.

<Comparative example 1>
FIG. 6(a) shows the result of inkjet printing of image data obtained by quaternary quantization at 600 dpi using the conventional error diffusion method. In FIG. 6A, the texture information is not sufficiently reproduced because the quantization error accompanying multi-level quantization is spatially diffused without considering the structure of the texture. That is, in a partially enlarged image, for example, vertical lines cannot be reproduced. Also, the printing process resulted in most of the texture being crushed by the ink.

<Example 1>
FIG. 6(b) shows the result of inkjet printing of multilevel quantized image data obtained by four-level quantization at 600 dpi using the image processing apparatus and image processing method of the present invention shown in FIGS. 1 and 2. . In the same manner as in the case of using C, M, Y, and K inks described above, four values were determined for each ink. The error diffusion unit 26 performs quantization of all pixels in order by raster scanning or vector scanning, pixel by pixel, by diffusing the error based on this threshold parameter. Quantization was performed in the same manner for each ink. In FIG. 6B according to the present invention, for example, the texture information of the input grayscale image is quantized and expressed in a manner that emphasizes it so that the vertical lines of the partially enlarged image are clearly expressed. It is now possible to obtain print results that preserve texture information even after the printing process.

As described above, according to the present invention, without changing the printing paper, ink, head, and other hardware, compared to the printing results of the conventional error diffusion method, the present invention can improve the texture and The advantage of improving texture was confirmed.

Furthermore, the conventional error diffusion method is also used as a technique for expressing low brightness contrast in high brightness displays. Also in the field of display technology, emphasizing texture within an image is an important issue. Therefore, if the quantization technology of the present invention is applied to a display, it is possible to realize a display such as a television that can be applied to emphasize texture.

Note that although the specific configuration of the texture detection section is not shown in the above, for example, the edge strength of an image is extracted using a differential filter. The extracted edges with high intensity include noise in the image and outlines of objects, so in order to detect texture information, it is necessary to remove them from the extracted edges. It is possible to detect only texture information by assuming that the case where the density variance value in the local region of each pixel is large is noise or the contour of an object, and multiplying the edge strength by the reciprocal of the noise or the contour of the object.

Although vertical lines, transparency (brightness information), edge information, or contour information have been described as texture information, other texture information may be used, such as texture information of an object, unevenness information related to a surface condition, etc.

Moreover, frequency subband information and the like can be considered as existing texture information used for texture enhancement processing, but other texture information may also be used.

Furthermore, in the image processing method using the image processing device 10B (FIGS. 3 and 4), the process of determining the amount of correction and controlling based on the amount of correction is repeated, but the number of times it is repeated is as follows. The larger the number of times, the better the quantization result can be obtained, but the smaller the number of times, the smaller the amount of calculation becomes. Further, a configuration in which the process is not repeated may be used, and in that case, the amount of calculation can be minimized.

INDUSTRIAL APPLICABILITY The present invention can be used industrially as an image processing device, an image processing method, and a program that can be applied to enhance texture in images such as photographs. The present invention can be suitably used in commercial printing presses, home printers, displays, televisions, digital cameras, image processing software, and the like.

10A, 10B: Image processing device, 12a, 12b: Roll, 14: Web-like printing base material, 16a, 16b, 16c, 16d: Inkjet head, 18: Head control unit, 20, 30: Multi-value input image data, 22 , 32: Texture detection unit, 24: Error distribution unit, 26, 42: Error diffusion unit, 28, 44: Multi-value quantized output image data, 30: Multi-value input image data, 34: Correction amount calculation unit, 36: Difference calculation unit, 38: Visual processing unit, 40: Multivalued image correction unit, 50: Inkjet printer.

Claims (7)

An image processing device that converts multi-value input image data into multi-value quantized output image data,
A texture detection unit detects texture information of the multi-value input image data, and a gradation conversion process is performed on the multi-value input image data so as to emphasize the texture information, thereby converting the multi-value input image data into the an error diffusion unit that converts into multi-value quantized output image data;
a difference calculation unit that generates a difference image between the multi-value input image data and the multi-value quantized image data quantized by the error diffusion unit; a visual processing unit that is a visual characteristic filter that inputs the difference image; comprising a modification amount calculation section that calculates a modification amount so as to preserve texture information, and a multivalued image modification section that modifies the multivalued input image data based on the modification amount,
An image processing device characterized in that the error diffusion section generates multi-value quantized output image data based on the output of the multi-value image correction section. The image processing apparatus according to claim 1, wherein the texture information is information on a texture portion, and the information on the texture portion is surface unevenness information, surface pattern information, brightness information, edge information, or contour information. . An image processing method for converting multi-value input image data into multi-value quantized output image data using the image processing device according to claim 1 or 2, comprising :
a texture detection step of detecting texture information of the multivalued input image data with a texture detection unit;
converting the multi-value input image data into the multi-value quantized output image data in an error diffusion unit by performing tone conversion processing on the multi-value input image data so as to emphasize the texture information; An image processing method comprising: An image processing method for converting multi-value input image data into multi-value quantized output image data using the image processing device according to claim 1 or 2, comprising:
Using the texture information of the multivalued input image data,
converting the multi-value input image data into the multi-value quantized output image data in an error diffusion unit by performing tone conversion processing on the multi-value input image data so as to emphasize the texture information; An image processing method comprising: A step of generating a difference image between the multi-value input image data and the multi-value quantized image data quantized by the error diffusion section in a difference calculation section, and inputting the difference image to a visual processing section that is a visual characteristic filter. a step of calculating a modification amount in a modification amount calculation section so as to save the texture information; a step of modifying the multi-value input image data in a multi-value image modification section based on the modification amount; 5. The image processing method according to claim 3, further comprising the step of generating the multi-value quantized output image data in the error diffusion section based on the output of the multi-value image correction section. 6. The image processing method according to claim 5 , wherein the step of calculating the correction amount in the correction amount calculating section is repeated. A program for causing a computer to function as each means of the image processing apparatus according to claim 1 or 2 .

Images