JP4027300B2 - Image processing method, image processing apparatus, image forming apparatus, computer program, and recording medium - Google Patents

Description

  The present invention relates to an image processing method, an image processing apparatus, an image forming apparatus, a computer program, and a recording medium that reduce the number of gradations of image data.

  For example, when a gradation image including a halftone is formed in a pseudo manner using a binary output image forming apparatus, it is necessary to perform a binarization process that reduces the number of gradations to two while considering gradation reproducibility. There is. As a binarization processing method, various methods such as a method of simply comparing with a threshold, a dither method, or an error diffusion method are conventionally used.

On the other hand, the first base coefficient set in the frequency domain generated by the forward transform (for example, discrete cosine transform) of the image data and the second frequency domain corresponding to the predetermined halftone texture frequency domain. There is a method of generating compressed image data by combining a base coefficient set and quantizing and entropy encoding the combined base coefficients (see, for example, Patent Document 1).
Japanese Patent Laid-Open No. 2002-10085

  FIG. 26 shows an example of a dither matrix used in the dither method. However, in order to simplify the explanation, a binarization dither method will be described as an example, but the basic idea is the same even with multiple values. In the dither method, each pixel value of 4 × 4 units of input image data of 16 gradations is used by using a 4 × 4 dither matrix in which threshold values of, for example, 0 to 15 are set according to positions as shown in FIG. Is compared with a threshold value to determine on / off of each pixel. When the input image data has 256 gradations, a value obtained by multiplying each value in FIG. 26 by 16 is used as the threshold value.

  However, in the dither method, since binarization processing is performed using a dither matrix having the same pattern, a regular periodic pattern is likely to occur in the binarized image, and a dither-specific repetitive texture is generated. There is.

  The error diffusion method is a method in which binarization is performed while distributing errors (hereinafter referred to as quantization errors) generated when binarizing each pixel of the original image to surrounding pixels that have not yet been binarized. is there. If the pixel to be binarized is the target pixel, the quantization error of the target pixel is weighted according to the relative position from the target pixel and then before the binarization process positioned around the target pixel. Is added to each pixel value.

FIG. 27 shows an example of a weighting coefficient matrix used in the error diffusion method. In the example of FIG. 27, a 2 × 3 weighting coefficient matrix including the pixel of interest (I _X , I _Y ) with the horizontal direction (processing direction) as the X direction and the vertical direction as the Y direction is shown. The weighting coefficient matrix indicates the weighting coefficient of each relative position (lower left neighbor, lower neighbor, lower right neighbor, right neighbor) with reference to the target pixel (I _X , I _Y ). For example, the target pixel (I _X , I _Y ) is compared with a threshold value. If the target pixel (I _X , I _Y ) is larger than the threshold value, the target pixel (I _X , I _Y ) is turned on. To. Next, the difference (quantization error) between the determined on / off pixel value and the pixel value of the target pixel (I _X , I _Y ) is calculated based on the weighting coefficient matrix before the surrounding binarization process. Distribute to pixels. However, the target pixel (I _X, I _Y) because it is quantized left of the pixel _{(I X-1, I Y} ) is the pixel of interest (I _X, I _Y) prior to the quantization error Not allocated.

For example, if the quantization error is Err, the pixel (I _{X + 1} , I _Y ) on the right side and the pixel (I _{X + 1} , I _{Y + 1} ) on the lower right side of the pixel of interest (I _X , I _Y ). ), Lower adjacent pixels (I _X , I _{Y + 1} ), and lower left adjacent pixels (I _X−1 , I _{Y + 1} ) are Err × (7/16) and Err × (1/16), respectively. , Err × (5/16), Err × (3/16) are allocated.

  In the error diffusion method, the quantization error is distributed to the surrounding unprocessed pixels based on the weighting coefficient matrix, so that the binarized image has less image quality compared to the dither method, such as a moire pattern is less likely to appear. Has the advantage of being superior.

  However, since the error is diffused based on the same matrix for each pixel, there is a problem that a worm (a portion where dots are partially connected) occurs in a highlight portion or the like. Measures such as adding noise to the threshold have been taken (for example, Shigeaki Kakutani, “Threshold Operation Method in Error Diffusion Method”, Journal of Electrophotographic Society, 1998, Vol. 37, No. 2, p. 186-192). reference).

  Recently, along with improvement in performance of image forming apparatuses such as ink jet printers, there are many image forming apparatuses that enable multi-value output such as ternary output or quaternary output instead of binary output. In the forming apparatus, for example, multilevel processing such as multilevel error diffusion is performed. The principle of the multi-value processing is basically the same as the binarization processing, but the input image data is quantized using two or more threshold values and binarized in that it outputs three-value or more image data. Is different.

  For example, in the case of performing quaternarization processing on 256-gradation image data having a density of 0 to 255, the output values (values quantized by the threshold processing) are set to 0, 85, 171, and 255, and the threshold is set to 42. , 128, 214. In this case, the output value is determined by sequentially comparing the pixel value of the target pixel and the three threshold values. For example, when the output value is smaller than the threshold value 42, it is determined as 0, otherwise it is determined as 85 when the output value is smaller than the threshold value 128, and otherwise, the output value is determined as 171 or 255 by comparison with the threshold value 214. .

  However, a multi-valued image such as a quaternary image has a problem that a dot pattern becomes uniform in an intermediate density region, and a tone gap (a phenomenon in which gradation changes discontinuously) occurs. For example, when the output value is 0, 85, 171 or 255, a tone gap is likely to occur near the intermediate density portion 85, 171 due to concentration of the same density. A tone gap always occurs as long as quantization processing is performed with a plurality of threshold values. For example, even in the case of binary output, a tone gap occurs near 0 or 255, but in the case of multi-level output, there are intermediate density output values that are easily noticeable by humans, so the tone gap in the intermediate density portion Is easy to stand out. It is difficult to completely prevent the tone gap in the intermediate density portion by an algorithm that appropriately selects a plurality of threshold values used in the intermediate density region and performs quantization processing.

  Further, in the method described in Japanese Patent Laid-Open No. 2002-10085, data (frequency component) converted into a predetermined halftone frequency region is used, but predetermined halftone data is used. Therefore, there is a problem that the same texture as the error diffusion method or dither method described above occurs. That is, the method disclosed in Japanese Patent Laid-Open No. 2002-10085 merely performs halftone processing using the same method as the above-described conventional method in the frequency domain.

  The present invention has been made in view of such circumstances, and by changing, for example, a high frequency component of image data, the problem of a tone gap generated near the output value of the image data is improved, and a high quality binary value is obtained. An object of the present invention is to provide an image processing method, an image processing apparatus, an image forming apparatus, a computer program, and a recording medium capable of generating an image or a quaternary image.

  In addition, the present invention suppresses the occurrence of textures generated by the dither method and worms generated by the error diffusion method by changing, for example, high-frequency components of the image data. It is another object of the present invention to provide an image processing method, an image processing apparatus, an image forming apparatus, a computer program, and a recording medium that can generate a quaternary image.

  In addition, the present invention provides an image processing method, an image processing apparatus, an image forming apparatus, a computer program, and a recording medium that can reduce the processing time without reducing the image quality by omitting the quantization process and the inverse quantization process. For other purposes.

  The present invention also provides an image processing apparatus that can improve the problem of tone gap and the like without degrading the image quality by changing the high frequency component without changing the low frequency component that occupies the main part of the image. Another object is to provide an image forming apparatus.

  The present invention also provides an image processing apparatus capable of generating a binary image or a quaternary image having smooth gradation reproducibility by changing the number of frequency components to be changed according to the magnitude of the DC component, and Another object is to provide an image forming apparatus.

An image processing method according to the present invention includes a step of converting image data into a spatial frequency component, a step of quantizing the converted spatial frequency component, and a spatial frequency in a predetermined frequency region with respect to the quantized spatial frequency component. A step of performing a change process for changing the component, a step of inversely quantizing the spatial frequency component on which the change process has been performed, a step of inversely transforming the spatially quantized spatial frequency component into image data, and an inverse transform Reducing the number of gradations of the image data based on a threshold, detecting a DC component included in the spatial frequency component, and determining a number of changes according to the detected DC component, in the step of performing a changing process for changing the spatial frequency components in a predetermined frequency range, the spatial frequency components of the number corresponding to the number of changes that the determined is changed And features.

The image processing method according to the present invention includes a step of converting image data into a spatial frequency component, a step of changing a spatial frequency component in a predetermined frequency region with respect to the converted spatial frequency component, and a changing process. A step of inversely converting the performed spatial frequency component into image data, a step of reducing the number of gradations of the inversely converted image data based on a threshold value, a step of detecting a direct current component included in the spatial frequency component, Determining the number of changes according to the detected DC component, and performing a change process for changing the spatial frequency component of the predetermined frequency region, wherein the number of spatial frequencies according to the determined number of changes The component is changed .

An image processing apparatus according to the present invention includes a frequency conversion unit that converts image data into a spatial frequency component, a quantization unit that quantizes the spatial frequency component converted by the frequency conversion unit, and a quantization unit that performs quantization. A changing unit that performs a changing process for changing the spatial frequency component of the predetermined frequency region, and a dequantizing unit that dequantizes the spatial frequency component subjected to the changing process by the changing unit, Inverse frequency conversion means for inversely converting the spatial frequency component inversely quantized by the inverse quantization means into image data, and reducing the number of gradations of the image data inversely converted by the inverse frequency conversion means based on a threshold value includes a threshold processing means, detecting means for detecting a direct current component included in the spatial frequency component, and determining means for determining a number of changes which the detecting means corresponding to the DC component detected, said changing means, said determination Characterized in that it is configured to change the number of spatial frequency components corresponding to the number of changes that stage is determined.

An image processing apparatus according to the present invention includes a frequency conversion unit that converts image data into a spatial frequency component, and a change process that changes a spatial frequency component in a predetermined frequency region with respect to the spatial frequency component converted by the frequency conversion unit. A changing means for performing, a quantizing means for quantizing the spatial frequency component subjected to the changing process by the changing means, an inverse quantizing means for dequantizing the spatial frequency component quantized by the quantizing means, Inverse frequency conversion means for inversely converting the spatial frequency component inversely quantized by the inverse quantization means into image data, and reducing the number of gradations of the image data inversely converted by the inverse frequency conversion means based on a threshold value includes a threshold processing means, detecting means for detecting a direct current component included in the spatial frequency component, and determining means for determining a number of changes which the detecting means corresponding to the DC component detected, said changing means, said determination Characterized in that it is configured to change the number of spatial frequency components corresponding to the number of changes that stage is determined.

An image processing apparatus according to the present invention includes a frequency conversion unit that converts image data into a spatial frequency component, and a change process that changes a spatial frequency component in a predetermined frequency region with respect to the spatial frequency component converted by the frequency conversion unit. A changing means to perform, an inverse frequency converting means for inversely converting the spatial frequency component subjected to the changing process by the changing means into image data, and the number of gradations of the image data inversely converted by the inverse frequency converting means as a threshold value. A threshold processing means for reducing the threshold based on the detection means, a detection means for detecting a DC component included in the spatial frequency component, and a determination means for determining the number of changes according to the DC component detected by the detection means. The number of spatial frequency components corresponding to the number of changes determined by the determining means is changed .

  An image forming apparatus according to the present invention includes the image processing apparatus, and is configured to perform image data forming processing with a reduced number of gradations using the image processing apparatus.

A computer program according to the present invention includes a procedure for causing a computer to convert image data into a spatial frequency component, a procedure for causing the computer to quantize the converted spatial frequency component, and a computer for converting the quantized spatial frequency component into a quantized spatial frequency component. On the other hand, a procedure for changing the spatial frequency component in the predetermined frequency region, a procedure for causing the computer to dequantize the spatial frequency component after the change processing, and a computer for causing the computer to inversely quantize the spatial frequency component. A procedure for inversely transforming the component into image data, a procedure for causing the computer to reduce the number of gradations of the inversely transformed image data based on a threshold, and a procedure for causing the computer to detect a DC component included in the spatial frequency component And a procedure for causing a computer to determine the number of changes according to the detected DC component, In the procedure to perform the changing process for changing the spatial frequency components, the spatial frequency components of the number corresponding to the number of changes that the determined is characterized in that it is changed.

A computer program according to the present invention causes a computer to convert image data into a spatial frequency component, and causes the computer to perform a change process for changing the spatial frequency component in a predetermined frequency region with respect to the converted spatial frequency component. A procedure for causing the computer to inversely convert the spatial frequency component subjected to the change processing to image data; a procedure for causing the computer to reduce the number of gradation levels of the inversely transformed image data based on a threshold ; and A changing process for changing the spatial frequency component of the predetermined frequency region, including a procedure for detecting a DC component included in the spatial frequency component and a procedure for causing the computer to determine the number of changes according to the detected DC component. in the procedure to perform, to characterized in that the spatial frequency components of the number corresponding to the number of changes that the determined is changed .

  A recording medium according to the present invention records the computer program.

  In the present invention, image data is converted into a spatial frequency component, a change process for changing the spatial frequency component in a predetermined frequency region is performed on the converted spatial frequency component, and the changed spatial frequency component is quantized. And inversely quantizing the quantized spatial frequency component, and inversely transforming the inversely quantized spatial frequency component into image data. The number of gradations of the inversely converted image data is reduced by, for example, quaternization using a threshold value. The image data with the reduced number of gradations is formed on, for example, a recording sheet. The changing unit performs, for example, a process for changing a spatial frequency component other than the low frequency region. Since the spatial frequency component in the low frequency region represents a rough portion of the image data, the change is easily noticeable to the human eye, and the image quality is degraded. By changing the spatial frequency component other than the low frequency region, the change becomes difficult to be noticed by humans. By changing the spatial frequency components other than the low frequency region and performing dot control of the image data, it is possible to suppress the occurrence of tone gaps, textures, worms, and the like. Further, since the number of gradations is simply reduced as compared with the threshold value, it is possible to improve the problem of texture or worm that has occurred in the dither method or error diffusion method.

  In the present invention, the image data is converted into a spatial frequency component, the converted spatial frequency component is quantized, and a change process for changing the spatial frequency component in a predetermined frequency region is performed on the quantized spatial frequency component, The spatial frequency component subjected to the change process is inversely quantized, and the inversely quantized spatial frequency component is inversely converted into image data. The number of gradations of the inversely converted image data is reduced by, for example, quaternization using a threshold value. The image data with the reduced number of gradations is formed on, for example, a recording sheet. The changing unit performs, for example, a process for changing a spatial frequency component other than the low frequency region. By changing the spatial frequency component other than the low frequency region, the change becomes difficult to be noticed by humans. By changing the spatial frequency components other than the low frequency region and performing dot control of the image data, it is possible to suppress the occurrence of tone gaps, textures, worms, and the like. Further, since the number of gradations is simply reduced as compared with the threshold value, it is possible to improve the problem of texture or worm that has occurred in the dither method or error diffusion method. Furthermore, since the spatial frequency component after quantization is changed, the numerical value handled by the change is smaller than that before quantization.

  In the present invention, image data is converted into a spatial frequency component, a change process for changing the spatial frequency component in a predetermined frequency region is performed on the converted spatial frequency component, and the spatial frequency component subjected to the change process is converted into an image. Convert back to data. The number of gradations of the inversely converted image data is reduced by, for example, quaternization using a threshold value. The image data with the reduced number of gradations is formed on, for example, a recording sheet. The changing unit performs, for example, a process for changing a spatial frequency component other than the low frequency region. By changing the spatial frequency component other than the low frequency region, the change becomes difficult to be noticed by humans. By changing the spatial frequency components other than the low frequency region and performing dot control of the image data, it is possible to suppress the occurrence of tone gaps, textures, worms, and the like. Further, since the number of gradations is simply reduced as compared with the threshold value, it is possible to improve the problem of texture or worm that has occurred in the dither method or error diffusion method. Furthermore, since the quantization process and the inverse quantization process are not performed, the processing time can be shortened. In addition, since no quantization error occurs, the image quality before threshold processing is improved, and the processing time can be shortened without reducing the image.

  In the present invention, the direct current component included in the spatial frequency component is detected by the detecting means, and the number of changes corresponding to the direct current component detected by the detecting means is determined by the determining means. The changing means performs processing for changing the number of spatial frequency components corresponding to the number of changes determined by the determining means. The number of spatial frequency components to be changed is controlled in accordance with a direct current component that is an average value of alternating current components included in the spatial frequency component.

  According to the present invention, by changing, for example, a high frequency component of image data, it is possible to improve a tone gap problem that occurs in the vicinity of an output value of the image data, and to generate a high-quality binary image or quaternary image. . In addition, by changing, for example, high-frequency components of image data, the generation of textures generated by the dither method and the occurrence of worms generated by the error diffusion method are suppressed. Can be generated.

  According to the present invention, since the number of gradations is simply reduced as compared with the threshold value, the problem of the texture or worm that has occurred in the dither method or the error diffusion method can be improved. The problem of texture or worm can be improved, and a high-quality binary image or quaternary image can be generated.

  According to the present invention, since the change process is performed on the spatial frequency component after quantization, the number of bits of the spatial frequency component handled in the change process is reduced, and the memory capacity can be reduced and the cost can be reduced. In addition, since the spatial frequency component after quantization has a small number of bits and is easy to handle, it is easy to adjust the dot control of the image data.

  According to the present invention, the processing time can be shortened by omitting the quantization process and the inverse quantization process. In addition, since no quantization error occurs, the image quality before the threshold processing is improved, and the processing time can be shortened without reducing the image.

  According to the present invention, since the low frequency component occupying the main part of the image is not changed and the high frequency component is changed, problems such as tone gap can be improved without degrading the image quality.

  According to the present invention, the number of output dots can be increased or decreased according to the image density by increasing or decreasing the number of spatial frequency components to be changed according to the magnitude of the direct current component, resulting in a smoother gradation. A halftone image having reproducibility (in which a tone gap is suppressed) can be generated.

  According to the present invention, the optimum change value can be set in advance by storing the change value of the spatial frequency component in advance. Even if binarization or quaternarization is performed with a threshold value by changing the spatial frequency component in the high frequency region by setting the change value in advance so that balanced dot reproducibility can be obtained, A toned binary image or quaternary image can be obtained. In addition, the change value can be determined easily and quickly without requiring complicated calculation processing.

Hereinafter, the present invention will be specifically described with reference to the drawings showing embodiments thereof.
(Embodiment 1)
FIG. 1 is a block diagram showing a configuration example of a gradation reproduction processing apparatus (image processing apparatus) 10 according to the present invention. The gradation reproduction processing device 10 generates output image data Po (X, Y) in which the number of gradations (for example, 256 gradations) of the input image data Pi (X, Y) is reduced to, for example, four values. It is. Here, the input image data Pi (X, Y) is on the Y-th line of the image data composed of pixels arranged in a two-dimensional matrix in the X direction (right direction) and the Y direction (down direction). This is pixel data at the Xth pixel position, and a two-dimensional image is constituted by a large number of input image data Pi (X, Y).

  The gradation reproduction processing apparatus 10 includes an image data storage unit 1 that stores input image data Pi (X, Y), and a frequency conversion that converts the input image data Pi (X, Y) into a spatial frequency component Qj (S, T). Part (frequency conversion means) 2, quantization part (quantization means) 3 for quantizing the spatial frequency component Qj (S, T), change for changing a part of the quantized spatial frequency component Qk (S, T) Section (changing means) 4, an inverse quantization section (inverse quantization means) 5 that inversely quantizes the spatial frequency component Ql (S, T) whose part has been changed, and an inversely quantized spatial frequency component Qm (S , T) is subjected to inverse frequency conversion (inverse frequency conversion means) 6, threshold value processing part (threshold processing means) 66 for performing threshold processing of image data Pn (X, Y) subjected to inverse frequency conversion, and Provided with a control unit (not shown) that controls each of the above units, and threshold processing Power image data Po (X, Y) and outputs a.

The image data storage unit 1 sequentially stores input image data Pi (X, Y) constituting a two-dimensional image. The input image data Pi (X, Y) is sequentially output to the frequency conversion unit 2 with 8 × 8 pixels as one block under the control of the control unit, for example. The frequency conversion unit 2 performs conversion to the frequency domain (frequency conversion) on the image data output in units of blocks. In this description, a discrete cosine transform (DCT) will be described as an example. Wherein the DCT is an input image A _ij, the output image B _ij, if a row of the input image A, the size of the column was M, and N, represented by the following formula.

  The frequency conversion unit 2 receives image data having 8 × 8 pixels as one block from the image data storage unit 1, performs DCT conversion, and DCT-converted spatial frequency components (hereinafter referred to as DCT coefficients) Qj (S, T) Is sent to the quantization unit 3. In this description, for a two-dimensional image, DCT conversion is performed in block units in the right direction (X direction) from the block including the upper left pixel, and finally the lowermost right is changed while changing the line in block units. DCT conversion is performed up to the final block including the pixels.

  The quantization unit 3 performs a quantization process on the DCT coefficient Qj (S, T) received from the frequency conversion unit 2. The DCT coefficient Qj (S, T) is divided by a constant threshold in the quantization unit 3. For example, all values are divided by 64. The changing unit 4 changes the quantized DCT coefficient Qk (S, T) in units of one block. The change is not performed for all the DCT coefficients in one block, but only for some of the 8 × 8 DCT coefficients in one block.

  2 and 3 are diagrams illustrating examples of regions (change regions) in which DCT coefficients are changed. The DCT coefficient obtained by DCT conversion of image data includes the uppermost DC component (hereinafter referred to as DC component) and the other AC component (hereinafter referred to as AC component). The DCT coefficient is not changed in the low frequency side (upper left) area including the DC component, and the DCT coefficient is changed in the high frequency side (lower right) area.

  In the example of FIG. 2 and FIG. 3, the frequency is changed in the first to fourth columns of the first row, the first to third columns of the second row, the first and second columns of the third row, and the first column of the fourth row. It is set as a non-change area that is not performed. In the example of FIG. 2, the 8th column of the 1st row, the 2nd row, the 7th to 8th columns, the 3rd row, the 6th to 8th columns, the 4th row, the 5th to 8th columns, the 5th row. The 4th to 8th columns, the 3rd to 8th columns of the 6th row, the 2nd to 8th columns of the 7th row, and the 1st to 8th columns of the 8th row are set as change regions for changing the frequency. In the example of FIG. 3, the third row, the third to eighth columns, the fourth row, the third to eighth columns, the fifth row, the third to eighth columns, the sixth row, the third to eighth columns, and the seventh row. The third to eighth columns of the eye and the third to eighth columns of the eighth row are changed areas. The change area may be set to an area other than the DC component. The change area is set in advance in, for example, the change unit 4 or a control unit (not shown).

  The change unit 4 replaces (changes) each DCT coefficient in the change area with, for example, 0, 1 or -1. FIG. 4 is a diagram illustrating an example of changing the DCT coefficient. 4 shows an example of changing the change area shown in FIG. 2, but the change is not limited to the change shown in FIG. For the replacement (change) of the DCT coefficient in the change area, for example, the DCT coefficient Qj (S, T) at the time of frequency conversion is compared with two threshold values Vth1, Vth2 (where Vth1> Vth2). Depending on the comparison result, 0, 1 or -1 can be substituted.

  FIG. 5 is a diagram illustrating an example of a change value of the DCT coefficient based on the comparison with the threshold value. FIG. 5 shows an example of a change value in the sixth to eighth columns of the seventh row of the DCT coefficient of the block shown in FIG. 4, and is −1 when the DCT coefficient Qj (S, T) is Vth2 or less. Replacement: When Vth1 or more, 1 is replaced, and when Vth1 to Vth2 is replaced, 0 is replaced. The change process of each block is rarely the same pattern, and is a random (irregular) change.

  The change unit 4 sends the DCT coefficient Ql (S, T) after the change process to the inverse quantization unit 5. The inverse quantization unit 5 performs inverse quantization on the DCT coefficient Ql (S, T) subjected to the change process by the change unit 4. In this description, all DCT coefficients in the block are multiplied by 64. The inverse frequency conversion unit 6 performs inverse frequency conversion on the DCT coefficient Qm (S, T) obtained by the inverse quantization unit 5 and converts it into density region data (image data). The two-dimensional inverse DCT transform performs the inverse transform of Equation 1. When a natural image is subjected to DCT transformation and quantization (1/64 times) is performed, the DCT coefficient is a value of about about −1 except for the DC component and the low frequency region, for example, as shown in FIG. The distribution has a value from 1 to around 1. When the data in the high frequency region is changed to 0, 1, or −1, it is possible to change the dot arrangement in the density space while minimizing the influence on the original image data.

The threshold processing unit 66 uses the plurality of threshold values for the density area data (image data) Pn (X, Y) received from the inverse frequency conversion unit 6, and multi-value density data (output image data) Po (X, Y). ). For example, in the case of quaternary output, using three threshold values,
If 0 <Pn (X, Y) ≦ 42, then Po (X, Y) = 0,
If 42 <Pn (X, Y) ≦ 127, Po (X, Y) = 85,
If 127 <Pn (X, Y) ≦ 212, Po (X, Y) = 171,
If 212 <Pn (X, Y) ≦ 255, Po (X, Y) = 255
Convert to

  In this way, the input image data Pi (X, Y) stored in the image data storage unit 1 is converted into the frequency domain and partly changed, then inversely converted, and finally by threshold processing, The output image data Po (X, Y) is obtained by reducing the number of gradations to 4 values for all pixels.

  When threshold processing is performed on a two-dimensional image, the threshold processing using the matrix of the same pattern in the prior art is not repeated, and noise is added by changing the spatial frequency component (DCT coefficient). Therefore, it is possible to improve the occurrence of worms, textures, tone gaps, and the like, which have been problems in the past.

(Embodiment 2)
FIG. 6 is a block diagram showing a configuration example of the gradation reproduction processing apparatus (image processing apparatus) 11 according to the present invention. Similar to the gradation reproduction processing unit 10 of the first embodiment shown in FIG. 1, the gradation reproduction processing device 11 is an image data storage unit 1, a frequency conversion unit 2, a quantization unit 3, a change unit 4, an inverse quantization. Unit 5, an inverse frequency conversion unit 6, and a threshold processing unit 66. However, the quantization unit 3 and the changing unit 4 are connected in the reverse order to that of the first embodiment (FIG. 1).

  In the present embodiment, the changing unit 4 performs a changing process of the changing region of the spatial frequency component (DCT coefficient) Qj (S, T) converted by the frequency converting unit 2. The change process is almost the same as in the first embodiment, but in this embodiment, since quantization by the quantization unit 3 has not yet been performed, the DCT coefficient in the change area is, for example, an output 4-value Is replaced with 0, -64 or 64. Next, the quantized DCT coefficient Qk (S, T) is quantized by, for example, dividing by a constant value 64 in the quantizing unit 3. Thereafter, the same processing as in the first embodiment is performed.

  Either the quantization of the DCT coefficient Qj (S, T) obtained by the frequency conversion unit 2 or the change of the change region may be performed first. However, when the quantization is performed first, the DCT coefficient becomes small. (In this description, 1/64), the numerical value handled in the change process is reduced, and the memory capacity or the burden of calculation processing can be reduced.

(Embodiment 3)
FIG. 7 is a block diagram showing a configuration example of the gradation reproduction processing apparatus (image processing apparatus) 12 according to the present invention. The gradation reproduction processing device 12 includes a spatial frequency component (DCT) quantized by the quantization unit 3 between the quantization unit 3 and the changing unit 4 of the gradation reproduction processing device 10 according to the first embodiment shown in FIG. DC component determination that determines the magnitude of the DC component of the coefficient (Qk (S, T)) and controls the number of 0 or 1 and −1 that the changing unit 4 changes based on the magnitude of the DC component Part 7 has been added. Here, the numbers of 1 and −1 are the same. For example, when the number of 1 and −1 is 2, the number of 1 is 2 and the number of −1 is 2. For example, when the number of 1 and −1 to be changed is 2, the two DCT coefficients in the change area are changed to 1, and the other two DCT coefficients are changed to −1.

  FIG. 8 is a diagram illustrating an example of control of the numbers 1 and −1 to be changed. In FIG. 8, quaternary output is taken as an example. As shown in FIG. 8, the DC component determination unit 7 controls the number of 1 and −1 to be changed in the range of 0 to 11, for example, according to the DC component (DCv). In the example of FIG. 8, the DC component (DCv) is divided into six ranges, and the number of 1 and −1 is increased or decreased in the range of 0 to 11 with respect to the increase of the DC component in each range.

  More specifically, when the DC component is 0, the numbers of 1 and −1 are 0. As the DC component increases, the numbers of 1 and −1 repeatedly increase or decrease, and the DC component has A = 31. At 875, the number of 1 and -1 is 0. When the DC component is A / 6, 3A / 6, or 5A / 6, the numbers 1 and −1 are the maximum value (11). In addition, when the DC component is 2A / 6 or 4A / 6, the numbers 1 and −1 are minimal values but do not become 0.

  When the numbers of 1 and −1 corresponding to the DC component are determined, the change area is replaced (changed) with the determined number of 1 and the same number of −1. For example, it is possible to store the map data shown in FIG. 8 in the DC component determination unit 7 and output the numbers 1 and −1 according to the DC component to the change unit 4. The change unit 4 replaces the output number of DCT coefficients with 1 or −1 (or adds 1 or −1) at random positions in the change region. In addition, the DC component determination unit 7 can obtain the numbers of 1 and −1 according to the DC component as a function. The process after the change is the same as in the first embodiment.

  The number of dots of the output image data Po (X, Y) varies depending on the ratio of 1, 0, and −1 arranged in the high frequency region. A gradation pattern having 256 gradations is not represented by a binary expression such as hitting a dot or not hitting, but 0 dot (not hitting a dot), 1 dot size, 2 dot size (2 dots represent 1 dot) Let us consider a case where the dot diameter is larger than 1 dot, and is expressed by 4 values of 3 dot size (3 dots represent 1 dot). Usually, the number of dots of 1 dot size is gradually increased from the low density to the high density from the white area where no dots are hit at all, and the 2 dot size is used after becoming saturated. In this way, a smooth gradation pattern is expressed while continuously switching up to 3 dot size.

  FIG. 9 is a diagram schematically illustrating gradation using dot sizes of three levels. When considering the relationship between the number of changed areas 1 and −1 of the DCT coefficient Ql (S, T) after the change and the number of dots of the output image data Po (X, Y), the numbers 1 and −1 are In the portion close to the minimum, the dot size is almost one type, and in the portion where the numbers 1 and −1 are close to the maximum, the two types of dot sizes are in a mixed state at substantially the same rate. When the density value changes from 0 to 255 (255 is white), the state is changed from the state of only about 3 dots to the state of mixing of 3 dots and 2 dots, and then from the state of only about 2 dots. The dot size and the 1 dot size are mixed, and the 1 dot size and the 0 dot size (no dot) are mixed from the state of only the 1 dot size. When such a state is viewed as the number of dots, the number of dots repeats increasing and decreasing alternately.

  In the present invention, the DCT coefficient in the change region (high frequency region) is changed, but increasing or decreasing the number of 0, 1, 1 in the DCT coefficient (AC component) controls the AC component. The number of dots will be controlled. Therefore, by controlling the number of DCT coefficients to be changed to a sine wave shape that gradually increases according to the DC component and then gradually decreases, as in the gradation pattern shown in FIG. Smooth gradation can be ensured. That is, it is possible to smoothly switch the dot sizes of different levels and to smooth the entire gradation.

  The DC component represents a region having a uniform image density (for example, a sky without a cloud), and the AC component represents a fine contrast of the image (for example, a distant tree leaf). The DCT transform is obtained by converting image data into frequency space information, and represents DC components, vertical brightness, horizontal brightness, and diagonal brightness (AC component). An AC component of “0” means that there is no corresponding frequency component. The AC component corresponds to the amplitude of each frequency (simple waveform of each cosine function), and indicates which frequency component is strong. Further, the DC coefficient indicates an average value of the entire waveform. From this, by setting 1 and −1 to the same number, it is possible to prevent the entire amplitude from being biased to either one.

(Embodiment 4)
FIG. 10 is a block diagram showing a configuration example of the gradation reproduction processing apparatus (image processing apparatus) 13 according to the present invention. The gradation reproduction processing device 13 is a table (based on the change of the DCT coefficient in the change region) between the changing unit 4 and the DC component determining unit 7 of the gradation reproduction processing device 12 of the third embodiment shown in FIG. An LUT unit 8 for storing an LUT (Look Up Table) is added.

  FIG. 11 is a diagram illustrating an example of changing the DCT coefficient using the LUT. However, the change area is the same as in FIG. The LUT unit 8 stores a plurality of LUTs in which change values (0, 1, −1) of the respective change portions in the change area are set. For example, a plurality of types of LUTs having different numbers of 0, 1 and −1 are stored. The LUT unit 8 selects an LUT corresponding to the number of 1 and −1 according to the DC component based on the determination result of the DC component determination unit 7, sends the selected LUT to the change unit 4, and the change unit 4 The DCT coefficient is changed based on the LUT. The DCT coefficient can be changed by replacing the change area of the quantized DCT coefficient Qk (S, T) with the value of the LUT change area, or adding the LUT value to the DCT coefficient Qk (S, T). It is.

  According to the present embodiment, generation of tone gaps, textures, and worms can be suppressed, and by using an LUT, an optimal change value in the change area can be set in advance, and the reproducibility of the output image can be improved. More improved. Even if the binarization or quaternarization is performed with the threshold by changing the DCT coefficient in the high frequency range by setting the change value in advance so that a balanced dot reproducibility can be obtained, the gradation as a whole A binary image or quaternary image can be obtained. Further, since the LUT is used, the optimum change value can be determined easily and quickly.

(Embodiment 5)
FIG. 12 is a block diagram showing a configuration example of the gradation reproduction processing apparatus (image processing apparatus) 14 according to the present invention. The gradation reproduction processing device 14 has substantially the same configuration as that of the gradation reproduction processing device 13 according to the fourth embodiment shown in FIG. 10, but instead of the LUT unit 8, an LUT based on blue noise (hereinafter referred to as blue noise). A blue noise LUT unit 9 in which the (LUT) is stored.

  Blue noise is pattern data having a spatial frequency that is difficult for the human eye to perceive. Human vision is almost insensitive above a certain spatial frequency, and the MTF (Modulation Transfer Function) of the visual system is known as a kind of low-pass filter (for example, Tsuyoshi Hamada, “High image quality in inkjet printers”). Technology ", Journal of the Imaging Society of Japan, 2001, Vol. 40, No. 3, p.239-243). Blue noise can be obtained by manipulating a pseudo-random pattern and generating a pattern in which the main component of the spatial frequency is distributed in a band equal to or higher than the cutoff frequency of the visual system MTF.

  By replacing such blue noise with a spatial frequency component and changing the DCT coefficient of the image data using the replaced spatial frequency component, it is possible to improve the dispersibility of dots in a highlight area where the dot density is sparse. it can. Moreover, the texture in the intermediate density portion can be improved.

  Blue noise is usually given as a 256 × 256 data matrix, and the data matrix is called a blue noise mask. FIG. 13 is a diagram illustrating an example of a blue noise mask used in the case of binary output. The data held in the blue noise LUT unit 9 is a data table obtained by performing DCT conversion on the blue noise data in FIG. 13 and dividing by a constant value so that the maximum absolute value is approximately 1. is there.

  FIG. 14 is a diagram illustrating an example of a data table (blue noise LUT) in which the blue noise mask illustrated in FIG. 13 is subjected to DCT conversion, further divided by a certain number, and optimized. The blue noise data optimized after DCT conversion has a value in the high frequency region. In this embodiment, one block of 8 × 8 is a basic unit, a total of 64 blocks of 8 blocks in the horizontal direction and 8 blocks in the vertical direction are set as one group, and the pattern of the same group is repeated. When the size of the blue noise mask is 64 × 64 (8 blocks × 8 blocks), a total of 65 blocks of LUT data including 64 blocks of blue noise LUT data and 8 × 8 all data “0”. Data is stored in the blue noise LUT unit 9.

  The DCT coefficient Qk (S, T) quantized by the quantizing unit 3 is processed in units of blocks, and the changing unit 4 performs a DCT coefficient changing process based on the blue noise LUT. The change is the same as the LUT described above (FIG. 10). However, when blue noise is used as the LUT data, the DC component determination unit 7 determines whether the DC coefficient is the minimum value or the maximum value. . FIG. 15 is a flowchart illustrating an example of a change processing procedure using a blue noise LUT. When the DC component determination unit 7 determines that the DC component is not the minimum value or the maximum value (step S200: NO), the blue noise LUT unit 9 selects the block selected in the same manner as the above-described LUT (FIG. 10), for example. The blue noise LUT is output to the changing unit 4 (step S202). When the DC component determination unit 7 determines that the DC component is the minimum value or the maximum value (step S200: YES), the blue noise LUT unit 9 has all the components “0” in the stored LUT. The LUT is sent to the changing unit 4 (step S204).

  In the case of blue noise, the low-frequency component is almost 0 as described above, and the high-frequency component is the main, so there is no need to separate the change area where the DCT coefficient is changed from the non-change area where the DCT coefficient is not changed. Blue noise data can be added to the DCT coefficient of one block consisting of the above. That is, in this case, the predetermined frequency region (change region) means the entire block. Further, since the blue noise data is added, the original data is not lost, and the deterioration of the image can be suppressed. The processing after changing the DCT coefficient is the same as in the fourth embodiment.

  Blue noise, which is difficult for humans to perceive, is converted into a spatial frequency component, and the spatial frequency component (DCT coefficient) in the high frequency region of the image data is changed using the converted spatial frequency component. It is possible to improve the dispersibility of the dots in the region or the texture in the intermediate density region. By changing the DCT coefficient in the high frequency region using normalized blue noise data, even if binarization or quaternarization is performed with a threshold value, the binary image or quaternary image with excellent gradation as a whole Etc. can be obtained.

  In the changing unit that performs the changing process for changing the spatial frequency component described above, each DCT coefficient in the changed region is replaced with 0, 1 or −1, or 0, 1 or −1 is added. It is not limited to an integer value such as 1, and a real value after the decimal point can be used. FIG. 16 is a diagram illustrating an example of a blue noise LUT using real values after the decimal point. By using a real value after the decimal point, it is possible to suppress the amount of information that is lost, so that it is possible to change the DCT coefficient in the change area more finely than the integer value, and smoother gradation reproduction An image can be obtained. Therefore, as a quantized spatial frequency component value (DCT coefficient) and a spatial frequency component value to be changed (changed value), real values after the decimal point are used, and blue noise data is used as the changing process to make the entire block. By adding to, an image with the best quality can be obtained.

(Embodiment 6)
FIG. 17 is a block diagram showing a configuration example of the gradation reproduction processing apparatus (image processing apparatus) 15 according to the present invention. The gradation reproduction processing device 15 includes an image data storage unit 1 that stores input image data Pi (X, Y), and a frequency conversion that converts the input image data Pi (X, Y) into a spatial frequency component Qj (S, T). Part (frequency conversion means) 2, a change part (change means) 44 for changing a part of the spatial frequency component Qj (S, T), and inverse frequency conversion of the changed spatial frequency component Qn (S, T). Inverse frequency conversion unit (inverse frequency conversion means) 6, threshold processing unit (threshold processing means) 66 for performing threshold processing of image data Pn1 (X, Y) subjected to inverse frequency conversion, and control of each of the above parts (not shown) A control unit is provided to output output image data Po (X, Y) subjected to threshold processing.

  The gradation reproduction processing device 15 of FIG. 17 has a configuration in which the quantization unit 3 and the inverse quantization unit 5 are omitted from the gradation reproduction processing device 10 of the first embodiment (FIG. 1). The basic operations of the image data storage unit 1, the frequency conversion unit 2, the change unit 44, the inverse frequency conversion unit 6, and the threshold processing unit 66 of the gradation reproduction processing device 15 are almost the same as those in the first embodiment. The number of gradations of the input image data Pi (X, Y), the number of gradations of the output image data Po (X, Y), and the like are assumed to be the same as in the first embodiment.

  The changing unit 44 changes the data Qj (S, T) frequency-converted by the frequency converting unit 2 in units of blocks such as 8 × 8 pixels. The change is not performed for all the DCT coefficients in one block, but only for some of the 8 × 8 DCT coefficients in one block. For example, although the DCT coefficient is changed in the change area shown in FIG. 2, the present invention is not limited to this, and the change area may be set to an area other than the DC component. The change area is set in advance in, for example, the change unit 44 or a control unit (not shown).

  However, since the quantization unit 3 is not provided in the present embodiment, the changing unit 44 replaces the changed region portion of the DCT coefficient Qj (S, T) subjected to frequency conversion with, for example, 0, 64, or −64. (change). FIG. 18 is a diagram illustrating an example of changing the DCT coefficient. For example, the replacement (change) of the DCT coefficient in the change area is performed by using two threshold values Vth1 ′ and Vth2 ′ (however, the DCT coefficient Qj (S, T) when frequency conversion is performed, as in the first embodiment. Vth1 ′> Vth2 ′) can be compared and replaced with 0, 64 or −64 depending on the comparison result.

  FIG. 19 is a diagram illustrating an example of a change value of the DCT coefficient based on the comparison with the threshold value. FIG. 19 shows an example of the changed values of the Yth column, the Y + 1th column, and the Y + 2th column of the XT row of the DCT coefficient of a certain block. When the DCT coefficient Qj (S, T) is Vth2 ′ or less, Replace with -64, replace with 64 if Vth1 'or more, replace with 0 if Vth1' to Vth2 '. The change process of each block is rarely the same pattern, and is a random (irregular) change. The changing unit 44 sends the DCT coefficient Qn (S, T) after the changing process to the inverse frequency converting unit 6.

The inverse frequency conversion unit 6 performs inverse frequency conversion on the DCT coefficient Qn (S, T) that has been changed by the changing unit 44, and converts it into density region data (image data). The two-dimensional inverse DCT transform performs the reverse process of the DCT transform. The threshold processing unit 66 uses the plurality of threshold values for the density region data (image data) Pn1 (X, Y) received from the inverse frequency conversion unit 6 to output multi-value density data (output). Image data is converted into Po (X, Y). For example, in the case of quaternary output, using three threshold values,
If 0 <Pn1 (X, Y) ≦ 42, then Po (X, Y) = 0,
If 42 <Pn1 (X, Y) ≦ 127, Po (X, Y) = 85,
If 127 <Pn1 (X, Y) ≦ 212, Po (X, Y) = 171,
If 212 <Pn1 (X, Y) ≦ 255, Po (X, Y) = 255
Convert to

  When threshold processing is performed on a two-dimensional image, the threshold processing using the matrix of the same pattern in the prior art is not repeated, and noise is added by changing the spatial frequency component (DCT coefficient). Therefore, it is possible to suppress the occurrence of worms, textures, tone gaps, and the like, which have been problems in the past. Further, since the quantization process and the inverse quantization process (bit shift process in two steps) are omitted, the processing time can be shortened. Since the quantization process and the inverse quantization process are omitted, no quantization error occurs, and the image quality before the threshold process is improved, so that the image quality of the output image data Po (X, Y) is not deteriorated. You can save time.

  The example in which the quantization unit 3 and the inverse quantization unit 5 are omitted from the gradation reproduction processing device 10 according to the first embodiment has been described above. However, the gradation reproduction processing device according to the third embodiment including the DC component determination unit 7. 12, the quantization unit 3 and the inverse quantization unit 5 are omitted, or the quantization unit 3 and the inverse quantization unit 5 are removed from the gradation reproduction processing device 13 according to the fourth embodiment including the DC component determination unit 7 and the LUT unit 8. It is also possible to omit the quantization unit 3 and the inverse quantization unit 5 from the gradation reproduction processing device 14 of the fifth embodiment that includes the DC component determination unit 7 and the blue noise LUT unit 9.

  FIG. 20 is a block diagram showing an example of the configuration of the gradation reproduction processing device (image processing device) 16 according to the present invention. The gradation reproduction processing device 16 is obtained by omitting the quantization unit 3 and the inverse quantization unit 5 from the gradation reproduction processing device 13 of the fourth embodiment (FIG. 10). Since the quantization unit and the inverse quantization unit are not provided, the DC component determination unit 77 is connected to the frequency conversion unit 2. The DC component determination unit 77 receives the DC component of the DCT coefficient Qj (S, T) from the frequency conversion unit 2 and determines 0 or the number of 64 and −64. The LUT unit 88 stores, for example, an LUT in which 1 and −1 of the LUT shown in FIG. 11 are changed to 64 and −64. The changing unit 44 changes (replaces or adds) the DCT coefficient Qj (S, T) received from the frequency converting unit 2 based on the LUT received from the LUT unit 88.

  Here, by adding the normalized blue noise data including the decimal point to the high frequency component of the DCT coefficient, it is possible to obtain the best quality image, and by not performing the quantization process and the inverse quantization process, Since the processing time can be shortened without degrading the image quality, it can be said that the mode in which normalized blue noise data including a decimal point is stored in the LUT unit (blue noise LUT unit) 88 of FIG. 20 is the best mode.

(Embodiment 7)
FIG. 21 is a block diagram showing a configuration example of the image forming apparatus 70 according to the present invention. In this description, the image forming apparatus 70 operates as a digital color copying machine. The image forming apparatus 70 includes a color image input device 30, a color image processing device 31, a color image output device 32, and an operation panel 33. Although not shown, a CPU (Central Processing Unit) that controls each device in the image forming apparatus 70 is provided.

  The color image input device 30 includes, for example, a CCD (Charge Coupled Device), and a reflected light image from a document is read by the CCD, and RGB (R: red, G: green, B: blue) analog signals are generated. Is done. The generated RGB analog signal is sent to the color image processing device 31.

  The color image processing apparatus 31 includes an A / D (analog / digital) conversion unit 311, a shading correction unit 312, an input tone correction unit 313, a region separation processing unit 314, a color correction unit 315, a black generation and under color removal unit 316, A spatial filter processing unit 317, an output tone correction unit 318, a tone reproduction processing unit 319, and a control unit (not shown) for controlling each unit are provided. The gradation reproduction processing unit 319 performs the same processing as the gradation reproduction processing devices (image processing devices) 10 to 16 of the first to sixth embodiments described above.

  The color image processing device 31 converts the RGB analog signal received from the color image input device 30 into an RGB digital signal, performs various image processing such as correction processing, and CMYK (C: cyan, M: magenta, Y: yellow). , K: black) digital color signal is generated, and the number of gradations of the generated CMYK digital signal is reduced to binary or quaternary. The binarized or quaternarized output image data is temporarily stored in a storage unit (not shown), and is output to the color image output device 32 at a predetermined timing.

  The A / D conversion unit 311 receives the RGB analog signal from the color image input device 30, converts the received RGB analog signal into an RGB digital signal, and sends the RGB digital signal to the shading correction unit 312. The shading correction unit 312 performs processing for removing various distortions generated in the illumination system, the imaging system, and the imaging system of the color image input device 30 on the RGB digital signal received from the A / D conversion unit 311. This is sent to the input tone correction unit 313. The input tone correction unit 313 adjusts the color balance for the RGB digital signal (RGB reflectance signal) received from the shading correction unit 312 and is processed by the image processing system employed in the color image processing device 31. It is converted into a density signal that can be easily processed and sent to the region separation processing unit 314.

  The region separation processing unit 314 separates each pixel in the image of the RGB digital signal received from the input tone correction unit 313 into one of a character region, a halftone dot region, and a photograph region, and based on the separation result, the pixel Is output to the color correction unit 315, the black generation and under color removal unit 316, the spatial filter processing unit 317, and the gradation reproduction processing unit 319. The RGB digital signal received from the input tone correction unit 313 is sent to the color correction unit 315 as it is.

  The color correction unit 315 converts the RGB digital signal sent from the input tone correction unit 313 into a CMY (C: cyan, M: magenta, Y: yellow) signal in order to perform color reproduction faithfully, After the process of removing color turbidity based on the spectral characteristics of the CMY color material including the unnecessary absorption component is performed, it is sent to the black generation and under color removal unit 316. The black generation and under color removal unit 316 performs black generation to generate a black signal (K signal) from the three color signals (C signal, M signal, and Y signal) of the CYM signal received from the color correction unit 315, A new CMY signal is generated by subtracting the K signal obtained by black generation from the CMY signal of CMY, and a four-color signal of CMYK (CMYK signal) is sent to the spatial filter processing unit 317.

As a general black generation process, there is a method of generating black using skeleton black. In this method, the input / output characteristic of the skeleton curve is y = f (x), the input data is C, M, Y, the output data is C ′, M ′, Y ′, K ′, UCR (Under Color If the removal rate is α (0 <α <1),
K ′ = f {min (C, M, Y)}
C ′ = C−αK ′
M ′ = M−αK ′
Y ′ = Y−αK ′
It is represented by

  The spatial filter processing unit 317 performs spatial filter processing using a digital filter on the image of the CMYK signal received from the black generation and under color removal unit 316 and corrects the spatial frequency characteristics to correct blur or Perform processing to improve graininess degradation. The output tone correction unit 318 performs output tone correction processing and the like, and the tone reproduction processing unit 319 performs predetermined processing on the image data of the CMYK signal based on the region identification signal.

  For example, a region separated as a character by the region separation processing unit 314 is emphasized at a high frequency by sharp enhancement processing included in the spatial filter processing performed by the spatial filter processing unit 317 in order to improve the reproducibility of black characters or color characters. Increase the amount. Further, the gradation reproduction processing unit 319 performs high-resolution binarization or multilevel conversion suitable for high-frequency reproduction.

  Further, with respect to the regions separated as halftone dots by the region separation processing unit 314, the spatial filter processing unit 317 performs low-pass filter processing for removing the input halftone component. The output tone correction unit 318 performs output tone correction processing for converting a signal such as a density signal into a halftone dot area ratio that is a characteristic value of the color image output device 32, and the tone reproduction processing unit 319 performs processing. Then, gradation reproduction processing (halftone generation) is performed so that the image is finally separated into pixels and binarized or multi-valued so that each gradation can be reproduced.

  Further, with respect to the region separated into photographs by the region separation processing unit 314, the gradation reproduction processing unit 319 performs binarization or multi-value processing that places importance on gradation reproducibility.

  The CMYK signal (image data) binarized or multivalued by the gradation reproduction processing unit 319 is sent to the color image output device 32. The color image output device 32 is a device that forms an image on a recording medium such as paper based on the CMYK signal received from the color image processing device 31. For example, an electrophotographic or inkjet color image output device can be used.

  The operation panel 33 is an input means for an operator to input an instruction by a key operation or the like. An operator instruction is output as a control signal from the operation panel 33 to the color image input device 30, the color image processing device 31, and the color image output device 32. According to an operator's instruction, an original image is read by the color image input device 30, and after data processing by the color image processing device 31, an image is formed on a recording medium by the color image output device 32, thereby functioning as a digital color copying machine. The above processing is controlled by a CPU (not shown).

(Embodiment 8)
FIG. 22 is a block diagram showing a configuration example of the image forming system 71 according to the present invention. The image forming system 71 includes a computer 40 and a printer 41. The printer 41 may be a digital multifunction machine having a copy function and a facsimile function in addition to the printer function. Further, the printer 41 performs electrophotographic or inkjet image formation.

  The image data is input to the computer 40 from a scanner or a digital camera, for example, and stored in a storage device (not shown). The image data input to the computer 40 can be processed and edited by executing various application programs. The computer 40 includes a color correction unit 45 that performs color correction processing of output image data, and a gradation reproduction processing unit that performs threshold processing for reducing the number of gradations (for example, 256 gradations) of the output image data to two values or four values. 46 and a printer language translation unit 47 that converts output image data into a printer language. The color correction unit 45 also performs black generation and under color removal processing. The gradation reproduction processing unit 46 corresponds to the gradation reproduction processing devices (image processing devices) 10 to 16 of the first to sixth embodiments. The data converted into the printer language by the printer language translation unit 47 is output to the printer 41 via the communication port 44 (RS232C, LAN, etc.).

  FIG. 23 is a block diagram illustrating a configuration example of a computer. The computer 40 includes a CPU (Central Processing Unit) 51, a RAM (Random Access Memory) 52 such as a DRAM, a hard disk drive (hereinafter abbreviated as a hard disk) 53, and an external storage unit such as a flexible disk drive or a CD-ROM drive. 54 and a communication port 44 for controlling communication with the printer 41 or the like. The computer 40 includes an input unit 55 such as a keyboard or a mouse, and a display unit 56 such as a display device.

  CPU51 controls each part 52-56 and 44 mentioned above. In addition, the CPU 51 stores the program or data received from the input unit 55 or the communication port 44 or the program or data read from the hard disk 53 or the external storage unit 54 in the RAM 52, and executes the program stored in the RAM 52 or the data Various processes such as computation are performed, and various process results or temporary data used for various processes are stored in the RAM 52. Data such as calculation results stored in the RAM 52 is stored in the hard disk 53 or output from the display unit 56 or the communication port 44 by the CPU 51.

  The CPU 51 includes the color correction unit 45, the gradation reproduction processing unit (for example, the frequency conversion unit 2, the quantization unit 3, the change unit 4, the inverse quantization unit 5, the inverse frequency conversion unit 6, the threshold processing unit illustrated in FIG. 66) and the printer language translation unit 47. The hard disk 53 operates as the image data storage unit 1 that stores image data.

  The computer program recorded on the recording medium 59 such as a CD-ROM is read by the external storage unit 54, stored in the hard disk 53 or the RAM 52, and executed by the CPU 51, whereby the CPU 51 can be operated as each unit described above. . It is also possible to receive a computer program from another device via the communication port 44 connected to a LAN or the like and store it in the hard disk 53 or the RAM 52.

  The recording medium 59 may be a storage medium that can carry a program and can be read directly or indirectly by a computer. For example, it may be a semiconductor element such as a ROM or a flash memory, a magnetic storage medium such as a flexible disk, a hard disk, an MD or a magnetic tape, or an optical storage medium such as a CD-ROM, MO or DVD. The reading method is not limited.

  FIG. 24 is a flowchart illustrating an example of the gradation reproduction processing procedure. However, a description will be given assuming that a four-value output image is obtained by performing frequency conversion processing by DCT. Further, it is assumed that the input image data Pi (X, Y) is stored in the hard disk 53. The image data stored in the hard disk 53 is read into the RAM 52 with 8 × 8 pixels as one block under the control of the CPU 51 (step S101). The read image data is DCT transformed by the CPU 51 (step S102), and the DCT coefficient Qj (S, T) after the DCT processing is stored in the RAM 52. Next, the CPU 51 divides all DCT coefficients in the block by a constant value (for example, 64) to perform quantization processing (step S103), and stores the DCT coefficients Qk (S, T) after division in the RAM 52. To do.

  Of the quantized (divided) DCT coefficients, the DCT coefficient (AC component) in the changed region is changed (replaced or added) to 0, 1, or −1 by the CPU 51 (step S104), and the changed DCT coefficient Ql. (S, T) is stored in the RAM 52. The DCT coefficient can be changed in the same manner as in the above-described embodiments. For example, the LUT is stored in the hard disk 53, the CPU 51 reads the LUT according to the DC component, and uses the read LUT. Thus, the DCT coefficient can be changed. Further, the change can replace the DCT coefficient based on the LUT or add the value of the LUT to the DCT coefficient.

  All the DCT coefficients Ql (S, T) in the block after the change are inversely quantized (for example, 64 times) by the CPU 51 (step S105) and stored in the RAM 52. The inversely quantized DCT coefficient Qm (S, T) is subjected to inverse DCT conversion from frequency domain data to density domain data (image data) by the CPU 51 (step S106), and converted image data Pn (X, Y). Is stored in the RAM 52.

  The CPU 51 performs threshold processing for converting the inversely converted image data Pn (X, Y) into a quaternary image (output image data) Po (X, Y) using a plurality of threshold values (step S107), and the RAM 52 or Store in the hard disk 53. Through the above-described S101 to S107, the threshold processing for one block of image data Pi (X, Y) is completed. Thereafter, the same operation is repeated until the threshold processing is completed for all blocks (step S108: NO).

  When the threshold processing for all blocks is completed (step S108: YES), the output image data Po (X, Y) is converted into the printer language by the CPU 51 and transmitted from the communication port 44 to the printer 41.

(Embodiment 9)
In the eighth embodiment, the gradation reproduction process can be performed without the quantization process (S103) and the inverse quantization process (S105). In this case, the CPU 51 of the computer 40 operates as the frequency converting unit 2, the changing unit 44, the inverse frequency converting unit 6, and the threshold processing unit 66 shown in FIG. In this case, the recording medium 59 stores a program for causing the computer 40 to perform frequency conversion, change, inverse frequency conversion, and threshold processing.

  FIG. 25 is a flowchart illustrating an example of the gradation reproduction processing procedure. However, a description will be given assuming that a four-value output image is obtained by performing frequency conversion processing by DCT. Further, it is assumed that the input image data Pi (X, Y) is stored in the hard disk 53. The image data stored in the hard disk 53 is read into the RAM 52 with 8 × 8 pixels as one block under the control of the CPU 51 (step S201). The read image data is DCT transformed by the CPU 51 (step S202), and the DCT coefficient Qj (S, T) after the DCT transformation is stored in the RAM 52.

  Next, in the DCT coefficient Qj, the DCT coefficient (AC component) in the changed region is changed (replaced or added) to 0, 64, or -64 by the CPU 51 (step S203), and the changed DCT coefficient Qn (S, T) is stored in the RAM 52. The DCT coefficient can be changed in the same manner as in the above-described embodiments. For example, the LUT is stored in the hard disk 53, the CPU 51 reads the LUT according to the DC component, and uses the read LUT. Thus, the DCT coefficient can be changed. In addition, the DCT coefficient can be replaced based on the LUT, or the value of the LUT can be added to the DCT coefficient. However, the value of the LUT is not 0, 1 or -1, but is 0, 64 or -64, for example.

  The changed DCT coefficient Qn (S, T) is subjected to inverse DCT conversion from frequency domain data to density domain data (image data) by the CPU 51 (step S204), and the converted image data Pn1 (X, Y) is stored in the RAM 52. Is remembered.

  The CPU 51 performs threshold processing for converting the image data Pn1 (X, Y) subjected to inverse frequency conversion into a quaternary image (output image data) Po (X, Y) using a plurality of threshold values (step S205). Alternatively, it is stored in the hard disk 53. By the above-described S201 to S205, the threshold processing for one block of image data Pi (X, Y) is completed. Thereafter, the same operation is repeated until threshold processing is completed for all the blocks (step S206: NO).

  In this way, the input image data Pi (X, Y) stored in the image data storage unit 1 is converted into a frequency domain and partially changed, and then subjected to inverse frequency conversion, and finally subjected to threshold processing. The output image data Po (X, Y) is obtained by reducing the number of gradations to 4 values for all pixels. When the threshold processing for all blocks is completed (step S206: YES), the output image data Po (X, Y) is converted into the printer language by the CPU 51 and transmitted from the communication port 44 to the printer 41.

  In each of the above-described embodiments, the number of gradations is reduced to four or the like by simple threshold processing, but it is also possible to use an error diffusion method or a dither method as the threshold processing. In this case, since the noise is added by changing the high frequency component of the image data before performing the error diffusion method or dither method, the worm, texture or rule is compared with the conventional error diffusion method or dither method. Generation of a typical pattern can be suppressed. However, the generation of worms or textures is suppressed by adding noise, but the image quality may be degraded by noise, so the high-frequency components should be changed using blue noise that is difficult for human eyes to recognize. Is preferred.

It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a figure which shows the example of the area | region (change area | region) which changes a DCT coefficient. It is a figure which shows the example of the area | region (change area | region) which changes a DCT coefficient. It is a figure which shows an example of a change of a DCT coefficient. It is a figure which shows an example of the change value of a DCT coefficient. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a figure which shows the example of control of the number of 1 and -1 which changes. It is a figure which shows the gradation using the dot size of three levels in simulation. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a figure which shows the example of a change of the DCT coefficient using LUT. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a figure which shows an example of a blue noise mask. It is a figure which shows an example of the data table (blue noise LUT) which optimized. It is a flowchart which shows the example of the change process sequence using a blue noise LUT. It is a figure which shows the example of the blue noise LUT using the real value below a decimal point. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. It is a figure which shows an example of a change of a DCT coefficient. It is a figure which shows an example of the change value of a DCT coefficient based on a comparison with a threshold value. It is a block diagram which shows the example of 1 structure of the gradation reproduction processing apparatus (image processing apparatus) which concerns on this invention. 1 is a block diagram illustrating a configuration example of an image forming apparatus according to the present invention. 1 is a block diagram illustrating a configuration example of an image forming system according to the present invention. It is a block diagram which shows the example of 1 structure of a computer. It is a flowchart which shows an example of a gradation reproduction process procedure. It is a flowchart which shows an example of a gradation reproduction process procedure. It is a figure which shows the example of a dither matrix. It is a figure which shows the example of a weighting coefficient matrix.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image data memory | storage part 2 Frequency conversion part 3 Quantization part 4, 44 Change part 5 Inverse quantization part 6 Inverse frequency conversion part 7, 77 DC component determination part (detection means, determination means)
8,88 LUT part (storage part)
9 Blue noise LUT part (storage part)
10, 11, 12, 13, 14, 15, 16 Tone reproduction processing device (image processing device)
31 Color Image Processing Device 40 Computer 46, 319 Gradation Reproduction Processing Unit 51 CPU
53 Hard Disk 66 Threshold Processing Unit 70 Image Forming Apparatus 71 Image Forming System

Claims (9)

Converting image data into spatial frequency components;
Quantizing the transformed spatial frequency component;
Performing a change process for changing a spatial frequency component of a predetermined frequency region with respect to the quantized spatial frequency component;
Dequantizing the spatial frequency component subjected to the change process;
Inversely transforming the inversely quantized spatial frequency component into image data;
Reducing the number of gradations of the inversely converted image data based on a threshold ;
Detecting a DC component included in the spatial frequency component;
Determining a number of changes according to the detected DC component;
Have
In the step of performing a changing process for changing a spatial frequency component in the predetermined frequency region, the number of spatial frequency components corresponding to the determined number of changes is changed . Converting image data into spatial frequency components;
Performing a change process for changing the spatial frequency component of the predetermined frequency region for the converted spatial frequency component;
A step of inversely converting the spatial frequency component subjected to the change processing into image data;
Reducing the number of gradations of the inversely converted image data based on a threshold ;
Detecting a DC component included in the spatial frequency component;
Determining a number of changes according to the detected DC component;
Have
In the step of performing a changing process for changing a spatial frequency component in the predetermined frequency region, the number of spatial frequency components corresponding to the determined number of changes is changed . Frequency conversion means for converting image data into spatial frequency components;
Quantization means for quantizing the spatial frequency component converted by the frequency conversion means;
Change means for performing change processing for changing a spatial frequency component in a predetermined frequency region with respect to the spatial frequency component quantized by the quantization means;
Inverse quantization means for inversely quantizing the spatial frequency component subjected to change processing by the change means;
Inverse frequency transforming means for inversely transforming the spatial frequency component inversely quantized by the inverse quantizing means into image data;
Threshold processing means for reducing the number of gradations of the image data inversely converted by the inverse frequency converting means based on the threshold ;
Detecting means for detecting a DC component included in the spatial frequency component;
Determining means for determining the number of changes according to the DC component detected by the detecting means;
With
The image processing apparatus , wherein the changing means is configured to change the number of spatial frequency components corresponding to the number of changes determined by the determining means . Frequency conversion means for converting image data into spatial frequency components;
Changing means for performing a changing process for changing a spatial frequency component in a predetermined frequency region with respect to the spatial frequency component converted by the frequency converting means;
Quantization means for quantizing the spatial frequency component subjected to change processing by the change means;
Inverse quantization means for inversely quantizing the spatial frequency component quantized by the quantization means;
Inverse frequency transforming means for inversely transforming the spatial frequency component inversely quantized by the inverse quantizing means into image data;
Threshold processing means for reducing the number of gradations of the image data inversely converted by the inverse frequency converting means based on the threshold ;
Detecting means for detecting a DC component included in the spatial frequency component;
Determining means for determining the number of changes according to the DC component detected by the detecting means;
With
The image processing apparatus , wherein the changing means is configured to change the number of spatial frequency components corresponding to the number of changes determined by the determining means . Frequency conversion means for converting image data into spatial frequency components;
Changing means for performing a changing process for changing a spatial frequency component in a predetermined frequency region with respect to the spatial frequency component converted by the frequency converting means;
Inverse frequency transforming means for inversely transforming the spatial frequency component subjected to the change processing by the changing means into image data;
Threshold processing means for reducing the number of gradations of the image data inversely converted by the inverse frequency converting means based on the threshold ;
Detecting means for detecting a DC component included in the spatial frequency component;
Determining means for determining the number of changes according to the DC component detected by the detecting means;
With
The image processing apparatus , wherein the changing means is configured to change the number of spatial frequency components corresponding to the number of changes determined by the determining means . An image forming apparatus comprising: the image processing apparatus according to claim 3 , wherein the image processing apparatus is configured to perform image data forming processing with a reduced number of gradations. apparatus. A procedure for causing a computer to convert image data into a spatial frequency component;
A procedure for causing the computer to quantize the transformed spatial frequency component;
A procedure for causing a computer to perform a changing process for changing a spatial frequency component in a predetermined frequency region with respect to a quantized spatial frequency component;
A procedure for causing the computer to dequantize the spatial frequency component that has undergone the change process;
A procedure for causing the computer to inversely transform the inversely quantized spatial frequency component into image data,
A procedure for causing the computer to reduce the number of gradations of the inversely converted image data based on a threshold value ;
A procedure for causing a computer to detect a DC component included in a spatial frequency component;
A procedure for causing the computer to determine the number of changes according to the detected DC component;
Including
A computer program characterized in that the number of spatial frequency components corresponding to the determined number of changes is changed in a procedure for performing a change process for changing a spatial frequency component in the predetermined frequency region . A procedure for causing a computer to convert image data into a spatial frequency component;
A procedure for causing a computer to perform a changing process for changing a spatial frequency component in a predetermined frequency region with respect to the converted spatial frequency component;
A procedure for causing the computer to reversely convert the spatial frequency component subjected to the change processing into image data,
A procedure for causing the computer to reduce the number of gradations of the inversely converted image data based on a threshold value ;
A procedure for causing a computer to detect a DC component included in a spatial frequency component;
A procedure for causing the computer to determine the number of changes according to the detected DC component;
Including
A computer program characterized in that the number of spatial frequency components corresponding to the determined number of changes is changed in a procedure for performing a change process for changing a spatial frequency component in the predetermined frequency region . 9. A computer-readable recording medium in which the computer program according to claim 7 or 8 is recorded.

Images