JPH09284429A - Picture evaluation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To execute picture evaluation matched with subjective evaluation by evaluating the quality of a picture at the time of displaying the picture inputted from a picture inputting equipment by giving filtering correction based on a human visual characteristic. SOLUTION: One patch in a gray chart inputted from the picture inputting equipment 12 is inputted to a picture evaluation device as a picture to be evaluated. Then picture data of the picture to be evaluated is matrix-transformed based on a characteristic parameter 15 by a picture evaluation processing part 17 to be transformed to tristimulus values. Next the tristimulus values are transformed to three sets of color systems in an opposite color space. Next the tristimulus values in the transformed opposite space are one-dimensionally and orthogonally transformed to obtain space frequency distribution. Then the obtained spatial frequency distribution is filter-corrected by the spatial frequency characteristic of human eyes suited to the human visual characteristic. The obtained spatial frequency distribution after filtering processing is orthogonally and inversely transformed to return to an opposite color reacting color.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image evaluation device obtained by an image input device such as an image scanner or a digital camera, and more particularly to an image quality evaluation technique that is consistent with human visual characteristics.

[0002]

2. Description of the Related Art Conventionally, an image to be evaluated (hereinafter,
An image to be evaluated) is a spatial frequency characteristic M of human vision.
The quality of the evaluated image is evaluated by matching the subjective evaluation that evaluates the sensitivity based on TF (Modulation Transfer Function) and the objective evaluation that evaluates the property of the image structure by the numerical value measured with a measuring instrument. The technique is widely applied to monochrome images.

Further, a technique for developing a target image into a color image is disclosed in, for example, Japanese Patent Laid-Open No. 5-284259. According to this technique, the image information is the brightness information,
It recognizes it as chromaticity information, calculates the spatial frequency distribution of this information, corrects it with the spatial frequency characteristics of the human visual system, and calculates the image evaluation value that represents the image quality. It is possible to perform image quality evaluation consistent with the physical evaluation.

In addition, in order to make up for the drawbacks of this technique,
Japanese Patent Application No. 7-313012 proposes a technique relating to an evaluation method in the case of having a peak at a specific frequency such as a halftone dot image in consideration of the fact that the visual MTF changes depending on the observation condition.

[0005]

However, the above-mentioned prior art algorithm is mainly intended for an image output by applying a constant input signal to an image output device such as an electrophotographic device, and is used in an image scanner, a digital camera or the like. There was no appropriate evaluation method for the evaluation of image input devices.

The present invention has been made in view of such conventional problems, and it is assumed that the image to be evaluated input by the image input device is observed on a display device such as a CRT monitor. It is an object of the present invention to provide an image evaluation device capable of performing accurate image quality evaluation that is consistent with subjective evaluation by human eyes.

[0007]

Therefore, according to the invention described in claim 1, the image information when the evaluated image input from the image input device is displayed on the image display device is converted into spatial frequency distribution information, The spatial frequency distribution information is filtered by a function that represents the spatial frequency characteristic of the human visual system according to the observation parameter, and then the image evaluation value is calculated from the image information obtained by performing the inverse transformation.

According to a second aspect of the present invention, image information when an image to be evaluated input from an image input device is displayed on an image display device is adapted to an ideal light source having a uniform spectral distribution, and the adapted image is adapted. The image information obtained by converting the information into the opposite color space is orthogonally transformed into the spatial frequency distribution information, and the spatial frequency distribution information is filtered by a function representing the spatial frequency characteristic of the human visual system according to the observation parameter, and then the inverse orthogonality is applied. The image evaluation value is calculated from the image information obtained by the conversion.

According to a third aspect of the present invention, there is provided an image evaluation apparatus for evaluating image quality when an image to be evaluated input from an image input device is displayed on an image display device, wherein the image input from the image input device is evaluated. Adaptation means for adapting information to an ideal light source having a uniform spectral distribution, opposite color space conversion means for converting the adapted image information into an opposite color space, and orthogonal conversion of opposite color response by the opposite color space conversion means Spatial frequency distribution generating means for generating spatial frequency distribution information, and processing the spatial frequency distribution information of the spatial frequency distribution means with a function representing the spatial frequency characteristic of the human visual system according to the observation parameter. An image evaluation value is calculated on the basis of the spatial frequency characteristic correction means for filtering correction by the An image evaluation value computing means that were to comprise a.

According to a fourth aspect of the present invention, the function representing the spatial frequency characteristic corresponds to an observation parameter including at least one of an observation distance, a luminance of an image to be evaluated, and ambient luminance of the image to be evaluated. And In the invention according to claim 5, the image evaluation value is calculated based on a statistical variation amount in the uniform color space by converting the image information obtained by the inverse orthogonal transform into the uniform color space. I did it.

In a sixth aspect of the present invention, the statistical variation amount is a standard deviation of image information obtained for each axis of the uniform color space. According to a seventh aspect of the invention, the image evaluation value is obtained by using at least one of the chromaticity of each primary color of the image display device and the white point, the gamma characteristic, and the resolution as a parameter.

According to an eighth aspect of the present invention, image information for a plurality of scanning lines is input at predetermined intervals in the direction perpendicular to the scanning line direction from the scanning start point of the image to be evaluated, and each scanning is performed. The one-dimensional orthogonal transformation and orthogonal inverse transformation are performed for each line. In the invention according to claim 9, the orthogonal transformation is a two-dimensional orthogonal transformation, and the orthogonal inverse transformation is a two-dimensional orthogonal inverse transformation.

According to a tenth aspect of the present invention, the image input device inputs the image information obtained by once photographing the image information on a film and developing the photographed film. In the invention described in claim 11, the image input device directly inputs image information to be observed.

In a twelfth aspect of the present invention, the image information input from the image input device is a gray chart having a plurality of uniform density areas from low density to high density.

[0015]

As described above, according to the invention described in claim 1, when the image to be evaluated obtained by the image input device is displayed on the image display device and observed on the image display device. The spatial frequency distribution obtained by orthogonally transforming the optical information of the image to be evaluated is filtered and corrected by the visual MTF that matches the observation conditions, thereby providing a highly reliable image evaluation that is consistent with the subjective evaluation. It can be carried out.

According to the second aspect of the present invention, after the optical information of the evaluated image obtained by observing the evaluated image obtained by the image input device on the CRT monitor is adapted to the ideal light source, By converting to the opposite color space, it is possible to obtain image information suitable for performing evaluation based on human visual characteristics while suppressing the occurrence of conversion error, and to obtain the converted image information by orthogonal transformation. By correcting the spatial frequency distribution and the visual MTF that matches the observation condition and filtering, the correction is performed with the optimum function according to the observation condition.
It is possible to perform highly reliable image quality evaluation that is consistent with subjective evaluation.

According to the third aspect of the present invention, the image to be evaluated inputted by the image input device is adapted to the ideal light source by the adapting means, and then converted to the opposite color space by the opposite color space converting means. Image information suitable for performing evaluation based on human visual characteristics while suppressing the occurrence of conversion error can be obtained. The converted image information is used to obtain spatial frequency distribution information by the spatial frequency distribution generation means, The frequency distribution information is corrected by an optimum function according to the observation condition by filtering and correcting the visual MTF that matches the observation condition by the spatial frequency characteristic correction means, so that the reliability is consistent with the subjective evaluation. High image quality evaluation can be performed.

According to the fourth aspect of the present invention, by setting the visual MTF according to various observation parameters, it is possible to use the optimum spatial frequency characteristic in accordance with the actual observation condition, and thus it is possible to obtain a higher accuracy. High image quality evaluation can be performed. According to the fifth aspect of the present invention, the noise component of the evaluated image can be quantitatively evaluated by performing the image quality evaluation based on the statistical variation amount in the uniform color space. A suitable image evaluation value can be obtained.

According to the sixth aspect of the invention, the image quality evaluation value of the image to be evaluated can be judged more realistically by setting the statistical variation amount to a standard deviation that is generally widely used. In addition, the statistical analysis of the image quality evaluation value can be performed more easily.
According to the invention of claim 7, by setting the visual MTF based on the characteristic of the image display device, the image quality of the evaluated image is evaluated without being affected by the image display characteristic of the image display device. Therefore, the reliability of image quality evaluation can be further improved.

According to the invention described in claim 8, since the image quality is evaluated for each scanning line, one of the orthogonal transformation processes is performed.
It is possible to reduce the calculation burden of the number of times, and 2
It is possible to evaluate a three-dimensional image, and it is possible to perform image quality evaluation without increasing costs such as adding new hardware. According to the invention described in claim 9, since the image information obtained in two dimensions is subjected to each two-dimensional orthogonal transformation processing, for example, when the hardware capable of high-speed processing is incorporated into the device, each orthogonal transformation is easily performed. The conversion can be done.

According to the tenth aspect of the invention, the image quality of various films can be evaluated by inputting the image information from the film on which the image information is recorded to an image input device such as a film scanner. Therefore, it is possible to further improve the image quality by performing noise reduction filtering processing or the like in advance when an image is input according to various film characteristics. In addition, it can contribute to the development of an image processing method for noise reduction and the like.

According to the eleventh aspect of the present invention, the image information to be observed is directly input by an image input device such as a digital camera, so that an intermediate image recording medium such as a film scanner is obtained. It is possible to evaluate the image quality of the image input device itself without considering the noise component of, and it is possible to contribute to the design of the image input device.

According to the twelfth aspect of the invention, the image quality evaluation for each density is systematically performed by inputting a gray chart having a plurality of uniform density areas from low density to high density to the image input device. The image quality evaluation result can be determined in a multidimensional manner.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The contents of the embodiments of the present invention will be described in detail below with reference to the drawings. In the present embodiment, the quality of an image input from an image input device is evaluated by applying filtering correction based on human visual characteristics to evaluate the quality of the image when the image input device is displayed. To do.

FIG. 1 is a block diagram showing the schematic arrangement of the image evaluation apparatus according to this embodiment. First, an image of the gray chart 11 having a pattern that gradually changes from high density to low density is taken under a standard light source (for example, D ₆₅ light source) by, for example, a commonly used film camera, and the film thus taken is taken. By the development processing, an image taken in a negative or positive state is reproduced. This reproduced image is used as an image input device 12 such as a film scanner.
The image is read while being digitized, and the captured image is converted into digital image data.

Next, the digitally converted image data is input to the image evaluation apparatus and displayed on the image display unit 13 such as a CRT monitor. In this embodiment, the image quality of the image displayed on the image display unit, that is, the image read by the image input device is evaluated based on human visual characteristics. In order to evaluate the image quality, the input parameter storage unit 14 has a gamma characteristic value of the image display unit 13 used for evaluation,
Chromaticity point for each color of Red, Green, Blue, White, and the characteristic parameter 15 such as the resolution of the image display unit, the observation distance when observing the image display unit, the screen average brightness,
Also, the observation parameter 16 such as the brightness of the peripheral portion of the screen is input.

Then, the image evaluation value processing unit 17 into which the image data output to the image display unit 13 and each parameter from the input parameter storage unit 14 is input calculates an image evaluation value, and the calculated image evaluation value is calculated. Outputs the image evaluation value
Output from 18. Here, in place of the film scanner for inputting the image of the film, a scanner such as a flatbed scanner or a drum scanner for reading an image of a printed matter or the like may be used, and a digital camera, a digital video camera or the like may be used as the image input device 12. Alternatively, a method of directly converting into a digital image may be used. Thereby, the input image of the image input device 12 itself can be evaluated.

Further, the image evaluation may be performed after properly correcting the gray balance of the captured image using the captured image of the captured gray chart. Furthermore, when a film camera is used, scratches and dust may adhere to the film surface. Therefore, by setting an appropriate threshold value and performing image processing such as binarization, scratches, dust, etc. The image quality evaluation may be performed after detecting the abnormalities in 1 above and eliminating these influences. By these processes, the image quality evaluation can be performed more accurately.

Next, an algorithm for image quality evaluation in the image evaluation processing section 17 will be described with reference to the flow chart shown in FIG. The image of the gray chart is input by the image input device 12 as described above (step 1, hereinafter referred to as S1). A uniform density value area is selectively extracted from the input image data, and the extracted one patch is used as an image to be evaluated. That is, when a gray chart is imaged, there are a plurality of patches having different density values, and the image quality is evaluated for each patch.

In the present embodiment, a case will be described in which a monochrome image in which a noise component under a D ₆₅ light source, which is often used as an evaluation target image, is superimposed is used. First, one patch of the gray chart input from the image input device is input to the image evaluation apparatus as an image to be evaluated. Then, in the image evaluation processing unit 17, the image data (R, G, B) of the evaluated image is subjected to matrix conversion from the digital value based on the characteristic parameter 15 stored in the input parameter storage unit 14, Convert to tristimulus values (X _D , Y _D , Z _D ) (S
2).

Specifically, the image display unit 13 is expressed by the equation (1).
Of the CRT monitor used as the image display unit 13 shown in FIG. 3 and the standardized screen brightness. A conversion matrix for converting into tristimulus values is obtained based on the characteristics.

[0032]

[Equation 1]

Here, X ₀ , Y ₀ , Z ₀ : tristimulus values at the white point of the monitor x _R , y _R : chromaticity x _G when the R component is 100% (G and B are 0%) , Y _G : Chromaticity when the G component is 100% (B and R are 0%) x _B , y _B : Chromaticity when the B component is 100% (R and G is 0%) The matrix converts digital image data (R, G, B) input from the image input device into tristimulus values (X _D , Y _D , Z _D ).

Next, the white point is changed under the E light source, which is an ideal light source in which each spectral component is uniform, using the adaptation model (S
3). The change of the white point under the E light source is a process necessary for further improving the accuracy of image quality evaluation in performing the process in the opposite color space described later. Equation (3) shows the white point conversion equation based on the von Kries model, which is widely used as an adaptation model. S _R '= (S _R / S _R0 ) ・ S _R1 S _G ' = (S _G / S _G0 ) ・ S _G1 (3) S _B '= (S _B / S _B0 ) ・ S _B1 where S _R0 , S _G0 , S _B0 : ternary color of the original reference white point S _R , S _G , S _B : current color S _R1 , S _G1 , S _B1 : reference white point S _R ', S _{G of the} reproduction destination ', S _B ': ternary color when adapted to the reproduction destination S _R , S _G , S _B in the equation (3) are tristimulus values X under the D ₆₅ light source
By substituting the white point of the D ₆₅ light source for _D , Y _D , Z _D , S _R0 , S _G0 , and S _B0 and the white point of the E light source to S _R1 , S _G1 , and S _B1 , respectively, a new light source under the E light source is obtained. Natristimulus value (X
_E , Y _E , Z _E ) is obtained. Although the von Kries model is used here, other models may be used if necessary.

Next, the tristimulus values (X _E , Y _E , Z _E ) under the E light source obtained by the equation (3) are converted into three sets (Red-Green, Yellow-Blue, White-) of the opposite color space. (Black) color system (S4). By the way, although there is luminance information, chromaticity information, etc. in the image information of a color image, conventionally, only the luminance information has been mainly used to evaluate an image. On the other hand, when evaluating a color image, there is a demand to evaluate the color balance and the like. In that case, the image is also evaluated by using the chromaticity information. However, since human visual characteristics are different for each color and each color component cannot be evaluated in a unified manner, human visual characteristics for each color are filtered and corrected for each color component of an image for evaluation. It has been empirically clarified that when the filtering correction is performed for each color component, a result that corresponds well to the human subjective evaluation can be obtained if the filtering correction is performed in the opposite color space. Therefore, also in the present embodiment, processing is performed using the opposite color space suitable for the filtering correction. Since the method of converting to the opposite color space is a well-known technique, it is omitted here.

Next, the stimulus values (R-G, Y-B, W-K) in the converted opposite color space are one-dimensionally orthogonally transformed in units of one scanning line of the image to be evaluated, and the resulting orthogonal transformation results are obtained. The spatial frequency distribution (R-G _DFT , Y-B _DFT , W-K _DFT ) is calculated by calculating the power spectrum (S
5). There are various methods for orthogonal transform, but here, discrete Fourier transform (DFT) is used.
Will be used. In addition, instead of DFT, FFT (Fast
It may be possible to increase the speed by using the Fourier Transform.

Then, the spatial frequency distribution obtained in S5 is subjected to filtering correction by the spatial frequency characteristic (MTF) of the human eye adapted to the human visual characteristic (SF).
6). The filtering correction method using this MTF will be described in detail below.

It is known that the spatial frequency characteristics of human vision differ depending on the observation conditions (observation parameters) such as the observation distance, the average luminance of the evaluated image at the time of observation, and the ambient luminance of the evaluated image. (Toyohiko Hatada "Display Condition and Visual MTF" No. 61 Oplus E, 1984). However, in the conventional image evaluation device, such an observation condition is not actually taken into consideration, and correction is performed only by one representative human visual system MTF, and when the observation condition is different, The subjective evaluation and the objective evaluation of the image quality evaluation cannot be matched.

FIG. 4 shows the MTF for each observation condition of the spatial sine wave pattern. FIG. 4 (a) shows the change in contrast sensitivity with respect to the observation distance when observing the image to be evaluated, using the observation distance as the observation parameter. In addition, FIG. 4B shows the change in contrast sensitivity with respect to the average luminance of the image to be evaluated when observing the image to be evaluated, using the average brightness as an observation parameter. Then, FIG. 4C shows the change in the contrast sensitivity with respect to the surrounding brightness of the evaluated image when the evaluated image is observed, using the peripheral brightness as the observation parameter.

Therefore, such observation distance, average brightness,
Numerical data of MTF corresponding to the observation parameter such as ambient brightness is tabulated and stored in the input parameter storage unit 14 in advance, and when the observation parameter indicating the actual observation condition is input from outside the apparatus, the image evaluation value is input. The processing unit 17 reads out numerical data corresponding to the observation parameter and the characteristic parameter of the image display unit 13 from the input parameter storage unit 14 and outputs the numerical data to the image evaluation value processing unit 17.

As a result, it is possible to accurately and easily select the appropriate MTF of the visual system corresponding to the actual observation conditions. In addition to the characteristic values shown in FIG. 4, the MTF of the visual system corresponding to each of the above observation parameters may be experimentally obtained by using the image evaluation device according to the present embodiment, and other documents etc. You may use the characteristics etc. which were described. Further, as the configuration of the input parameter storage unit 14, for example, a memory such as an EPROM, a storage device such as a magnetic or optical disk, and the like can be cited. By using these storage devices, random access of data or the like is possible. Then, the desired information can be retrieved quickly.

Then, in the image evaluation value processing unit 17, the numerical data of the spatial frequency distribution of the image to be evaluated (R-G _DFT , which is obtained by orthogonally transforming the numerical data of MTF corresponding to the observation conditions in S5) is obtained. Y-B _DFT , W-K _DFT ) is applied for filtering correction. As a result, the frequency components that humans are strongly sensitive to are emphasized more, while the frequency components that are less sensitive are attenuated.
Spatial frequency component sensitivity (R-
G _C , Y-B _C , W-K _C ) can be converted, and the MTF (filter) of an appropriate visual system corresponding to the observation condition is selected and the filtering correction is performed. Therefore, the effect and reliability of the correction process can be further improved. Can be increased.

The spatial frequency distribution after the filtering process thus obtained is subjected to orthogonal inverse transformation to return to the opposite color response value (S7). Again, S5
Similarly, by performing a discrete Fourier inverse transform DIFT, which is a kind of orthogonal inverse transform, in the opposite color space (R
-G ', Y-B', W-K '). Next, the opposite color response value (R-
G ′, Y−B ′, W−K ′) are tristimulus values (X _E ′,
By returning to Y _E ', Y _E ') (S8) and processing S3 in the reverse direction, the white point is changed to the original D ₆₅ light source (S9). As a result, new tristimulus values (X _D ', Y _D ', Y _D ') under the D ₆₅ light source are obtained.

Next, the tristimulus values (X _D ', Y _D ', Y _D ') under the D ₆₅ light source are converted into a uniform color space (S10). This uniform color system includes, for example, CIEL ^*
u ^* v ^* (L ^* u ^* v ^* ) and CIELAB (L ^* a ^* b)
^* ), Etc., but in the present embodiment, CIEL ^* u
A uniform color system of ^* v ^* will be used. Since the method of converting the tristimulus values into the uniform color space of CIEL ^* u ^* v ^* is a well-known technique, its explanation is omitted here.

Since the above-mentioned processing is processing for one scanning line of the image to be evaluated, the processing of S2 to S10 described above is repeated for all scanning lines of the image to be evaluated (S1).
1) Plot the processing results in the uniform color space. FIG. 5 shows a conceptual diagram of the uniform color space in which the results of each processing are plotted. As shown in FIG. 5, the coordinates in the uniform color space for each pixel are not all concentrated on one point, but exhibit a distribution with some variation. This degree of variation represents the noise component of the evaluated image.

For example, when an image of completely uniform density is input to the image input device, if the image input device is an ideal input means that does not cause noise to be superimposed, the coordinates of the image to be evaluated in the uniform color space will be one point. I will concentrate. Further, when there is a noise component such as unevenness in the input image, or when the noise component is superimposed on the input image by the image input device, the coordinates have variations.

Therefore, in order to quantify this variation,
For each L ^* , u ^* , v ^* axis, for example, a statistical amount L such as standard deviation
^* Noise, u ^* Noise, and v ^* Noise are calculated respectively.
Then, in order to match these statistics with human subjective evaluation, after multiplying by a coefficient experimentally obtained in advance, the total is obtained, and this is the image evaluation value of the evaluated image, that is, noise of the image input device. (Scanner Noise) (S12). (4)
The formula shows an example of a specific calculation formula of the image evaluation value. Scanner Noise = L ^* Noise + 0.852u ^* Noise + 0.323v ^* Noise (4) Here, when the evaluated image is a monochrome image, L ^*
The image evaluation value may be the statistic amount for the axis, that is, only L ^* Noise.

Then, the evaluation result by the image evaluation value processing unit 17 is output to the output unit 18 such as a CRT monitor or a printer.
FIG. 6 shows the results of evaluating three types of films (ISO400 negative, ISO100 negative, ISO100 reversal) using the evaluation method according to the present embodiment. This result is in good agreement with the result of human subjective evaluation. In addition, the evaluation result for each film is that the higher the sensitivity of the film is, the larger the image noise becomes. Therefore, by performing filtering processing to reduce the noise according to the film sensitivity at the time of image input or before image input, smoothness with less noise feeling It is possible to obtain a clear image.

In this embodiment, the orthogonal transformation and the inverse orthogonal transformation are performed one-dimensionally.
(Degital Signal Processor) or high-speed CPU
A two-dimensional MT is used as an MTF for filtering processing while replacing the two-dimensional orthogonal transformation and the two-dimensional inverse orthogonal transformation
By using F, the image quality can be evaluated easily and with an improved conversion processing speed.

Further, the algorithm applied to the image evaluation may be performed by a hardware method by executing a program previously installed in a memory such as a ROM, or may be performed by software. Further, these controls are performed based on a computer program stored in the image evaluation value processing unit. However, in combination with the sophistication of program technology or circuit design technology, how to control the whole by software control and hardware control It can be arbitrarily decided whether to divide it into two.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image evaluation apparatus according to this embodiment.

FIG. 2 is a flowchart showing an image evaluation algorithm in this embodiment.

FIG. 3 is a diagram showing a gamma characteristic of an evaluation CRT monitor.

FIG. 4 is a diagram showing visual spatial frequency characteristics with respect to various observation parameters.

FIG. 5 is a diagram showing a statistical variation amount in a uniform color space.

FIG. 6 is a diagram showing image quality evaluation results for various films.

[Explanation of symbols]

 11 Gray chart 12 Image input device 13 Image display unit 14 Input parameter storage unit 15 Image display unit Characteristic parameter 16 Observation parameter 17 Image evaluation value processing unit 18 Output unit

Claims (12)

[Claims] 1. An image information when an image to be evaluated input from an image input device is displayed on an image display device is converted into spatial frequency distribution information, and the spatial frequency distribution information of a human visual system according to an observation parameter. An image evaluation device characterized by calculating an image evaluation value from image information obtained by performing inverse transformation after filtering processing by a function representing spatial frequency characteristics. 2. Image information when an image to be evaluated input from an image input device is displayed on an image display device is adapted to an ideal light source having a uniform spectral distribution characteristic, and the adapted image information is stored in an opposite color space. It is obtained by orthogonally transforming the converted image information into spatial frequency distribution information, filtering the spatial frequency distribution information with a function representing the spatial frequency characteristic of the human visual system according to the observation parameter, and then performing inverse orthogonal transformation. An image evaluation device characterized by calculating an image evaluation value from image information. 3. An image evaluation apparatus for evaluating image quality when an image to be evaluated input from an image input device is displayed on an image display device, wherein image information input from the image input device has uniform spectral distribution characteristics. Adapting means for accommodating an ideal light source having: an opposite color space converting means for converting the adapted image information into an opposite color space; and spatial frequency distribution information by orthogonally converting the opposite color response by the opposite color space converting means. Spatial frequency distribution generation means for generating, for the spatial frequency distribution information of the spatial frequency distribution means,
Spatial frequency characteristic correction means for performing filtering correction by performing a calculation process with a function representing the spatial frequency characteristic of the human visual system according to the observation parameter, and image information obtained by inverse orthogonal transform of the output of the spatial frequency characteristic correction means And an image evaluation value calculation means for calculating an image evaluation value based on the image evaluation value. 4. The function representing the spatial frequency characteristic is in accordance with an observation parameter including at least one of an observation distance, a luminance of an evaluated image, and an ambient luminance of the evaluated image. Item 5. The image evaluation device according to any one of items 3. 5. The image evaluation value is calculated based on a statistical variation amount in the uniform color space by converting the image information obtained by the inverse orthogonal transform into a uniform color space. The image evaluation device according to any one of claims 2 to 4. 6. The statistical variation amount is a standard deviation of image information obtained for each axis of the uniform color space.
The image evaluation device described in 1. 7. The image evaluation value is obtained by using at least one of the chromaticity of each primary color of an image display device and the white point, gamma characteristic, and resolution as a parameter. The image evaluation device according to one. 8. Image information for a plurality of scanning lines is input at a predetermined interval in a direction perpendicular to a scanning line direction from a scanning start point of the image to be evaluated, and each scanning line is one-dimensional. The image evaluation device according to any one of claims 1 to 7, which is configured to perform orthogonal transformation and inverse orthogonal transformation. 9. The image evaluation apparatus according to claim 1, wherein the orthogonal transform is a two-dimensional orthogonal transform, and the orthogonal inverse transform is a two-dimensional orthogonal inverse transform. 10. The image input device, wherein the image information is once photographed on a film, and the image information obtained by developing the photographed film is inputted.
The image evaluation device according to any one of 1. 11. The image input device directly inputs image information to be observed.
The image evaluation device according to any one of 1. 12. The image according to claim 1, wherein the image information input from the image input device is a gray chart having a plurality of uniform density areas from low density to high density. Evaluation device.