JP5289222B2 - Image processing apparatus and method - Google Patents

Description

  The present invention relates to image processing for restoring image quality degradation caused by an image output device.

  When inputting / outputting image data using an image input / output device, there is a process for correcting deterioration of the frequency characteristics of the image data, such as a decrease in sharpness or blurring of the image by the input / output device (hereinafter referred to as restoration process). . In the restoration process, generally, image data converted into frequency space data using Fourier transform or the like is subjected to filtering by a restoration filter. For example, Patent Document 1 performs restoration processing using a Wiener filter that takes into account the influence of noise. Patent Document 2 introduces an adjustment parameter so that the user can adjust the degree of restoration when performing restoration processing using a filter.

  However, the input / output device may have a limit of restoration processing (hereinafter, restoration limit). The restoration limit when an image is printed by a printer will be described with reference to FIG. When an image having a luminance change shown in FIG. 1 (a) is printed, the contrast decreases as shown in FIG. 1 (b) due to the spatial frequency transfer function (MTF) characteristic of the printer.

  Therefore, restoration processing using a restoration filter (inverse filter) that is an inverse function of MTF is performed, and the luminance change of the image data supplied to the printer is corrected to the change shown in FIG. However, in the case of 8-bit image processing, it is necessary to round the image data to a range of values 0 to 255, and data outside the range is clipped. As a result, the luminance change of the image data output to the printer is a change indicated by a solid line in which the portion indicated by the broken line in FIG. During printing, deterioration due to the MTF characteristics of the printer occurs, and the luminance change of the output image is as shown in FIG. That is, although the contrast is improved as compared with FIG. 1B, the gradation change is degraded.

  Therefore, as shown in FIG. 1 (e), if restoration processing is performed in consideration of the maximum contrast at which clipping does not occur, an output image with good contrast and gradation changes can be obtained as shown in FIG. 1 (f). be able to. That is, this maximum contrast is an example of the restoration limit.

JP 2005-063000 JP JP 2007-183842 A

  An object of the present invention is to suppress image quality deterioration caused by an image output device in consideration of a restoration limit.

  The present invention has the following configuration as one means for achieving the above object.

  The image processing according to the present invention analyzes a frequency characteristic of an image output to an image output device, creates a restoration filter based on a spatial frequency transfer function of the image output device, and a restoration filter based on the frequency characteristic, A restoration limit is detected by comparing a spatial frequency transfer function and the frequency characteristic, and restoration processing for restoring image quality degradation caused by the image output device is performed using any of the restoration filters based on the restoration limit. It is characterized by applying to.

  Also, a restoration filter based on the spatial frequency transfer function and a restoration filter based on the visual transfer function of the image output device of the image output destination are created, and the spatial frequency transfer function and the visual transfer function are compared for each frequency. Setting the spatial frequency transfer function or the visual transfer function as a restoration limit, and using any one of the restoration filters based on the restoration limit, the restoration process for restoring the image quality degradation caused by the image output device It is applied to an image.

  According to the present invention, it is possible to suppress deterioration in image quality caused by the image output device in consideration of the restoration limit.

The figure explaining the restoration limit in the case of printing an image with a printer. FIG. 2 is a block diagram for explaining a configuration example of an image processing apparatus according to the first embodiment. 6 is a flowchart for explaining restoration processing of the image processing apparatus. The figure explaining the dependence of the frequency characteristic of a contrast fall. The figure which shows an example of an MTF characteristic. The figure explaining an example of the chart for MTF measurement. The figure explaining an example of the chart for MTF measurement. The figure explaining the example in which the frequency characteristic of an image deteriorates with the MTF characteristic of an image output apparatus. The figure explaining a restoration filter. FIG. 3 is a block diagram for explaining an example configuration of an image processing apparatus according to a second embodiment. 9 is a flowchart for explaining restoration processing of the image processing apparatus according to the second embodiment. The flowchart explaining the process of a restoration limit setting part in detail. The figure which shows a VTF characteristic.

  Hereinafter, image processing according to an embodiment of the present invention will be described in detail with reference to the drawings.

[Device configuration]
A configuration example of the image processing apparatus 100 according to the first embodiment will be described with reference to the block diagram of FIG.

  The image output device 102 is a printer that prints an image represented by input image data, such as an inkjet printer or an electrophotographic printer. The measurement device 103 is a device that measures the density distribution of the chart output from the image output device 102, such as a scanner or a digital camera. The image storage unit 106 is a storage device that stores various images such as image data to be printed by the image output device 102. A server device or the like connected to the network may be used instead of the storage device.

  In the image processing apparatus 100, the MTF characteristic calculation unit 14 calculates the spatial frequency transfer function (MTF) characteristic of the image output apparatus 102 based on the measurement data of the chart input from the measurement apparatus 103. The image input unit 17 reads image data stored in the image storage unit 106. The frequency characteristic analysis unit 18 performs Fourier transform on the image data read by the image input unit 17 and analyzes the frequency characteristic of the image data. The restoration filter creation unit 15 creates a restoration filter based on the MTF characteristics of the image output device 102 and the frequency characteristics of the image data.

  The restoration limit detection unit 19 detects the restoration limit of the image output device 102 based on the MTF characteristics of the image output device 102 and the frequency characteristics of the image data. The restoration filter setting unit 20 sets a restoration filter to be applied to the restoration process based on the restoration limit of the image output device 102 detected by the restoration limit detection unit 19.

  The restoration unit 21 performs a restoration process using the restoration filter set by the restoration filter setting unit 20 on the image data in the frequency space obtained by the Fourier transform of the frequency characteristic analysis unit 18. The inverse transform unit 22 inversely transforms the restored frequency space image data into real space image data by inverse Fourier transform. The image output unit 23 outputs the image data that has been subjected to the restoration process and converted back to the real space to the image output device 102 that is the output destination.

[Restore processing]
The restoration process of the image processing apparatus 100 will be described with reference to the flowchart of FIG.

  The MTF characteristic calculation unit 14 calculates the MTF characteristic of the image output apparatus 102 by inputting the measurement result of the MTF measurement chart printed by the image output apparatus 102 from the measurement apparatus 103, as will be described in detail later (S201). The restoration filter creation unit 15 calculates an inverse function of the MTF characteristic of the image output device 102 and an inverse function of the frequency characteristic of the original image, and creates a restoration filter (S202).

  Next, the image input unit 17 reads image data of an image to be printed (hereinafter referred to as an original image) stored in the image storage unit 106 (S203). The frequency characteristic analysis unit 18 performs Fourier transform on the image data of the original image and analyzes the frequency characteristic of the original image (S204). Although details will be described later, the restoration limit detection unit 19 detects the restoration limit of the image output apparatus 102 based on the MTF characteristics of the image output apparatus 102 and the frequency characteristics of the original image (S205).

  Next, as will be described in detail later, the restoration filter setting unit 20 sets a restoration filter to be applied to the restoration process based on the restoration limit of the image output device 102 detected by the restoration limit detection unit 19 (S206). The restoration unit 21 performs restoration processing using the set restoration filter on the image data in the frequency space of the original image obtained by the Fourier transform of the frequency characteristic analysis unit 18 (S207).

  Next, the inverse transform unit 22 performs inverse Fourier transform on the restored frequency space image data (hereinafter, image data of the restored image) (S208). The image output unit 23 outputs the image data of the restored image to the image output device 102 (S209), and causes the image output device 102 to print the restored image.

[MTF characteristic calculator]
In general, when an image is output by an image output device, the contrast decreases due to the characteristics of the image output device, and the degree of the decrease in contrast depends on the frequency characteristics of the input image. With reference to FIG. 4, the dependence of the frequency characteristics of the contrast reduction will be described.

  4 (a), 4 (c), and 4 (e) show sinusoidal luminance changes with the same amplitude but different frequencies. When image data indicating such a luminance change is input to an image output device, output results shown in FIGS. 4B, 4D, and 4F are obtained, respectively. That is, even in a sinusoidal luminance change having the same amplitude (same contrast), the degree of contrast reduction in the output image differs depending on the frequency.

  Therefore, when the ratio of the contrast of the input data and the contrast of the output image is plotted with the horizontal axis representing the resolution (frequency) and the vertical axis representing the contrast ratio, an MTF characteristic as shown in FIG. 5 is obtained.

  An example of the MTF measurement chart will be described with reference to FIGS. When measuring the MTF, a chart for measurement having a sinusoidal luminance change as shown in FIG. 6 and changing the frequency of the luminance change is printed on the image output device 102, and the image of the chart is measured by the measuring device 103. Get the data. The MTF characteristic calculation unit 14 calculates the contrast ratio from the contrast of the input data and the contrast of the image data of the chart for each frequency of the luminance change, and obtains the MTF characteristic of the image output device 102.

  FIG. 6 shows an example of a chart in which the luminance changes in the horizontal direction. Further, if a chart in which the luminance changes in the vertical direction as shown in FIG. 7 is used, MTF characteristics in the horizontal direction and the vertical direction can be obtained, and the diagonal direction is A two-dimensional MTF characteristic can be calculated by interpolation.

  If the MTF characteristics of the image output device depend not only on the frequency and two-dimensional direction, but also on the output position of the image, the brightness of the image, the amplitude of the image data, etc., a corresponding measurement chart is prepared. To do. For example, a chart showing sinusoidal luminance changes with the same frequency and different amplitudes may be printed, an MTF characteristic corresponding to the characteristic of the image data may be selected, and a restoration filter may be created. In other words, as long as it is possible to calculate the MTF characteristics of the image output device desired by the user, the measurement method and calculation method are not limited.

Restore filter control and settings
An example in which the frequency characteristics of an image deteriorate due to the MTF characteristics of the image output device will be described with reference to FIG. FIG. 8 (a) shows an example of frequency characteristics of the original image, and FIG. 8 (b) shows an example of MTF characteristics of the image output device. The frequency characteristic of the output image is a characteristic indicated by a solid line in FIG. 8C, which is obtained by multiplying the characteristic of FIG. 8A by the characteristic of FIG. 8B.

  The restoration filter will be described with reference to FIG. Ideally, an inverse function (that is, an inverse filter) of the MTF characteristic (FIG. 8 (b)) of the image output device is created as shown in FIG. 9 (a). Then, by applying the inverse function to the frequency characteristics (FIG. 8 (a)) of the original image, the MTF characteristics of the image output device can be canceled and the frequency characteristics of the original image and the output image can be matched.

  However, as shown in FIG. 9 (a), the absolute value of the inverse filter increases rapidly in a high frequency region where the MTF characteristic approaches zero. When the original image is restored using such an inverse filter, some pixels of the restored image deviate from the data range of the image output device (for example, 0 to 255 in the case of an 8-bit image). As a result, a process (clip) for rounding the pixel value to the data range of the image output device is required, and an output image having an expected frequency characteristic cannot be obtained.

  Therefore, the MTF characteristic indicating the maximum contrast that can be output by the image output device is used. That is, in the frequency range where the power of the original image does not exceed the MTF characteristics of the image output device, the power of the original image is set as a restoration target value. Conversely, in the frequency range where the power of the original image exceeds the MTF characteristic of the image output device, the value of the MTF characteristic is set as a target value for restoration.

That is, the restoration limit detection unit 19 detects the frequency νs at which the power of the original image exceeds the MTF characteristic of the image output device 102 as the restoration limit of the image output device 102. Then, the restoration filter setting unit 20 applies the inverse filter of the MTF characteristic G (ν) to the restoration processing of the frequency component whose frequency is below the restoration limit, and the power of the original image to the restoration processing of the frequency component whose frequency exceeds the restoration limit. Apply an inverse filter of F (ν). That is, the restoration filter is expressed by the following equation.
if (ν ≦ νs)
H (ν) = 1 / G (ν);
else
H (ν) = 1 / F (ν);… (1)
Where F (ν) is the power spectrum of the original image,
G (ν) is the MTF characteristic of the image output device,
ν is the spatial frequency (cycles / degree).

  FIG. 9 (b) shows the frequency characteristics of the restoration filter of Equation (1), and if the restored image obtained by applying this restoration filter to the original image is output by the image output device, the output of the frequency characteristics shown in FIG. 9 (c) An image is obtained. That is, an output image having the best frequency characteristic that can be output by the image output device can be obtained.

[Modification]
When the restoration process is performed by convolution of the real space filter, the restoration limit is detected and the restoration filter is set for each pixel of the original image. Thereby, it is possible to cope with a difference in restoration limit that is different for each pixel.

  In the above description, a single image output device 102 has been described. However, when a plurality of image output devices are operated in tandem, individual differences (manufacturing errors) of these image output devices are taken into consideration. That is, the MTF characteristic of the device having the worst characteristics is measured or estimated, a restoration filter is created so as not to exceed the restoration limit in any apparatus, and the restoration limit is detected.

  The image processing according to the second embodiment of the present invention will be described below. Note that the same reference numerals in the second embodiment denote the same parts as in the first embodiment, and a detailed description thereof will be omitted.

  A configuration example of the image processing apparatus 100 according to the second embodiment will be described with reference to the block diagram of FIG. The image processing apparatus 100 according to the second embodiment includes a Fourier transform unit 25 that only performs Fourier transform on the image data of the original image, instead of the frequency characteristic analysis unit 18 according to the first embodiment. In addition, a restoration limit setting unit 26 is provided instead of the restoration limit detection unit 19. Furthermore, a VTF characteristic storage unit 24 that stores human visual transfer function (VTF) characteristics is provided, and the restoration filter creation unit 15 creates a restoration filter based on the MTF characteristics and VTF characteristics of the image output device 102. . Other configurations are the same as those of the image processing apparatus 100 of the first embodiment.

  The restoration process of the image processing apparatus 100 according to the second embodiment will be described with reference to the flowchart of FIG. In the second embodiment, the Fourier transform unit 25 performs Fourier transform on the image data of the original image (S304), and the restoration limit setting unit 26 controls the restoration filter based on the MTF characteristic and the VTF characteristic of the image output device 102 (S305). This is different from the restoration process of the first embodiment. In step S202, the point that the restoration filter creation unit 15 creates the restoration filter does not change, but the difference is that the inverse function of the restoration filter is the MTF characteristic and the VTF characteristic of the image output device 102. Other processes are the same as those in the first embodiment.

  The processing of the restoration limit setting unit 26 will be described in detail with reference to the flowchart of FIG.

  The restoration limit setting unit 26 reads the MTF characteristic of the image output device 102 (S401), reads the VTF characteristic stored in the VTF characteristic storage unit 24 (S402), and sets the frequency of interest to a DC component (frequency ν = 0). (S403).

  Next, the restoration limit setting unit 26 compares the MTF characteristic and VTF characteristic of the frequency of interest (S404), and sets a characteristic having a small value as the restoration limit of the frequency (S405). Then, it is determined whether or not the restoration limit has been set for all frequencies for which the restoration limit is set (S406) .If there is an unset frequency, the target frequency is moved to the high frequency side (S407), and the process is stepped. Return to S404.

  In this way, the restoration limit setting unit 26 sets the frequency restoration limit corresponding to the frequency component output by the Fourier transform unit 25 to the MTF characteristic or the VTF characteristic. The restoration filter setting unit 20 sets a restoration filter to be applied to the restoration process according to the restoration limit set by the restoration limit setting unit 26. In other words, the restoration filter H (ν) = 1 / G (ν) is set for the restoration processing of the frequency component corresponding to the frequency of the MTF characteristic G (ν) whose restoration limit is the frequency, and the restoration limit is the frequency of the VTF characteristic V (ν). A restoration filter H (ν) = 1 / V (ν) is set in the restoration processing of the frequency component corresponding to. Therefore, the restoration unit 21 performs the restoration process using the restoration filter corresponding to the restoration limit set for each frequency by the restoration limit setting unit 26 on each frequency component.

[Restore limit setting]
Details of setting the restoration limit will be described. The VTF characteristic, which is the transfer function of human vision, can be approximated by the following equation.
if (ν ≧ 5)
VTF = 5.05e ^-0.138ν (1-e ^-0.1ν );
else
VTF = 1;… (2)

  Plotting the value of VTF obtained by equation (2) gives the characteristics shown in FIG. As shown in FIG. 13, the visual sensitivity decreases as the frequency increases. That is, humans cannot perceive a contrast smaller than the VTF characteristic. In other words, it is perceived that the image quality deterioration has been improved by performing the restoration process according to the VTF characteristics. Therefore, in determining the restoration limit, the MTF characteristic and the VTF characteristic of the image output apparatus 102 are compared, and the smaller value is set as the restoration limit.

[Modification]
In the second embodiment, the restoration process in the image output device has been described. However, the same restoration process can be performed in the image input device. That is, the MTF characteristic of the image input device is calculated, and a restoration filter is created from the MTF characteristic and the VTF characteristic. Then, the restoration limit is set to any one by comparing the MTF characteristic and the VTF characteristic, and restoration processing using a restoration filter corresponding to the set restoration limit is performed on the frequency component of the original image input from the image input device. Then, if the restored image is subjected to inverse Fourier transform, image deterioration caused by the image input device can be suppressed.

[Other Examples]
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, etc.) of the system or apparatus reads the program. It is a process to be executed.

Claims (9)

Analysis means for analyzing the frequency characteristics of the image output to the image output device;
A restoration filter based on a spatial frequency transfer function of the image output device, and a creation means for creating a restoration filter based on the frequency characteristics;
Detecting means for comparing the spatial frequency transfer function and the frequency characteristic to detect a restoration limit;
An image processing apparatus comprising: a restoration unit configured to perform restoration processing for restoring image quality deterioration caused by the image output device using any one of the restoration filters based on the restoration limit.   2. The image processing apparatus according to claim 1, wherein the detection unit detects a frequency at which a power of the image indicated by the frequency characteristic exceeds a value of the spatial frequency transfer function as the restoration limit.   3. The restoration means uses a restoration filter based on the spatial frequency transfer function at a frequency below the restoration limit, and uses a restoration filter based on the frequency characteristic at a frequency exceeding the restoration limit. The image processing apparatus described in 1. A creation filter for creating a restoration filter based on a spatial frequency transfer function of an image output device to which an image is output, and a restoration filter based on a visual transfer function;
For each frequency, the setting means for comparing the spatial frequency transfer function and the visual transfer function and setting the spatial frequency transfer function or the visual transfer function as a restoration limit;
An image processing apparatus comprising: a restoration unit configured to perform restoration processing for restoring image quality deterioration caused by the image output device using any one of the restoration filters based on the restoration limit.   5. The setting means, for each frequency, compares the value of the spatial frequency transfer function and the value of the visual transfer function, and sets the smaller value as the restoration limit. Image processing apparatus.   6. The image processing apparatus according to claim 1, further comprising a calculation unit that calculates the spatial frequency transfer function from an image output by the image output apparatus. Analyzing the frequency characteristics of the image output to the image output device,
Creating a restoration filter based on a spatial frequency transfer function of the image output device, and a restoration filter based on the frequency characteristic;
Comparing the spatial frequency transfer function and the frequency characteristic to detect a restoration limit;
An image processing method, wherein a restoration process for restoring image quality degradation caused by the image output device is performed on the image using any one of the restoration filters based on the restoration limit. Create a restoration filter based on the spatial frequency transfer function of the image output device of the image output destination and a restoration filter based on the visual transfer function,
For each frequency, compare the spatial frequency transfer function and the visual transfer function, and set the spatial frequency transfer function or the visual transfer function as a restoration limit,
An image processing method, wherein a restoration process for restoring image quality degradation caused by the image output device is performed on the image using any one of the restoration filters based on the restoration limit.   7. A program for controlling a computer device to function as each unit of the image processing device according to claim 1.