JPH11120325A - Image evaluation method, medium storing image evaluation program, and image evaluation apparatus - Google Patents

Abstract

(57) [Summary] [Problem] There is a possibility that bias will occur if only one evaluation criterion is used. SOLUTION: Computer 21 which is the center of image processing
Sample the pixel image data with different evaluation criteria in steps S120 and S140, and determine the weighting coefficient k in steps S192 to S196 based on the image evaluation option input in step S180. By using the weighted coefficient k and summing up the summation results in step S310 to generate a luminance distribution histogram, the image is evaluated based on the total summation result obtained by summing up a plurality of evaluation criteria.
Optimal image processing can be executed in 10 to S350.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image evaluation method for evaluating an image based on real image data such as a digital photographic image, a medium storing an image evaluation program, and an image evaluation apparatus.

[0002]

2. Description of the Related Art Various types of image processing are performed on image data of actual photographs such as digital photographic images. For example, image processing such as increasing contrast, correcting color tone, or correcting brightness. Normally, these image processings can be executed by a microcomputer, and an operator checks an image on a monitor and selects necessary image processing, or determines parameters of the image processing. That is, the operator determines the characteristics of the image and selects or executes various operations.

[0003]

In recent years, various image processing techniques have been proposed and have actually been effective. However, when it comes to how much processing is done with which technique, humans still have to be involved.
This is because it is not possible to determine where the digital image data to be subjected to image processing is important.

[0004] For example, in consideration of image processing for correcting brightness, it is assumed that automatic processing is performed in which, if the average of the entire screen is dark, the image is corrected to be bright, and if the average is bright, the image is corrected to be dark. Here, it is assumed that there is image data of a real image of a person image taken at night. It is assumed that although the background is almost completely dark, the person itself has been well photographed. If the image data of the actual photograph is automatically corrected, an attempt is made to correct the image data brightly because the background is completely dark, resulting in a daytime image.

In this case, if a human is involved, attention is paid only to the part of the human figure. Then, if the person image is dark, the correction is made slightly brighter, and conversely, if the image is too bright due to the effect of flash or the like, the correction is made darker.

In view of such a problem, the present applicant has proposed an invention in Japanese Patent Application No. xx to judge an important part in an image. In the present invention, it is considered that an original subject (object) should exist in a sharp portion of an image, and a pixel having a large degree of change is determined as an object by focusing on the degree of change of the image at each pixel. doing.

However, even if there is a sharp portion of the image, the area of the portion is small and image processing based on the background portion may be preferable in some cases, and there is a possibility that bias is caused by only one evaluation criterion. Further, the need to determine which evaluation criterion should be adopted still remains.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has an image evaluation method and an image evaluation method capable of flexibly responding to image evaluation as a prerequisite for image processing. It is an object to provide a medium on which an evaluation program is recorded and an image evaluation device.

[0009]

In order to achieve the above object, the invention according to claim 1 inputs real image data composed of pixels in a dot matrix form, and totals the image data of each pixel on a predetermined basis. An image evaluation method for evaluating an image on the basis of a totaling result, having a plurality of evaluation criteria for the totaling result, and summing the evaluation results based on each evaluation standard with a predetermined weight.

In the invention according to claim 1 configured as described above, as a premise of the evaluation method, image data of a real photograph consisting of pixels in a dot matrix is input, and the image data of each pixel is determined based on a predetermined standard. Aggregates and evaluates the images based on the aggregation results. Here, there are a plurality of evaluation criteria for the result of the aggregation, and the evaluation results based on the respective evaluation criteria are added together with a predetermined weight.

That is, there is an evaluation criterion suitable for evaluating an image with a sharp object of the image as an object such as a portrait, and an evaluation criterion suitable for evaluating the image with a background as an important object. There is also a plurality of evaluation criteria, which are executed in parallel while appropriately changing the weights of the respective evaluation criteria, and comprehensively evaluated.

The evaluation result may be any one that can be used to determine the characteristics of the image, and it is not necessary to obtain a result that specifically specifies the type of the image. For example, it also includes an index such as a histogram of luminance when an image is determined to be bright or dark, and it is not necessary to obtain a determination result such as a bright image or a dark image. Of course, other than light and dark, it may be an index for determining whether an image is sharp or an index for determining vividness.

In order to apply a plurality of evaluation criteria to the tabulation results, various methods that achieve substantially the same purpose can be adopted. For example, an example is that weighting is performed on all pixels using individual evaluation criteria and totalization is performed. However, since the processing amount increases if the totalization is performed for all the pixels, as an example suitable for such a situation, the invention according to claim 2 is a method according to claim 1, wherein When thinning out and summing up data according to the standard, the data is sampled based on a plurality of evaluation criteria and summed up, and the respective summation results are summed up with a predetermined weight.

[0014] In the invention according to claim 2 configured as described above, the image data is sampled as a precondition for counting, and a plurality of standards are eventually adopted by changing the standard for this sampling method. Further, by adjusting the weights of the respective totalization results and adding them together, the result is that the evaluation results based on a plurality of evaluation criteria are weighted and evaluated.

As a basic example of such an evaluation criterion, the invention according to claim 3 is based on claim 1 or claim 2.
In the image evaluation method described in any one of the above, one evaluation criterion is configured to uniformly sample and total.

According to the third aspect of the present invention, the image data is uniformly thinned out, but the entire image is captured. I can say.

On the other hand, the invention according to claim 4 is an example of an evaluation criterion applicable to both cases where the sampling method is adopted and cases where it is not adopted.
In one of the image evaluation methods described in any one of the above, one evaluation criterion is configured so that evaluation is made heavier for pixels having a large degree of change between each pixel and an adjacent pixel, and aggregation is performed.

In the invention according to claim 4 configured as described above, the degree of change between each pixel and an adjacent pixel is detected, and the pixels having a large degree of change are weighted and aggregated, and as a result, An evaluation criterion for evaluating a clear image portion having a large degree of change is adopted.

This evaluation criterion is evaluated with emphasis on a sharp part of the image, and it is needless to say that it is suitable for determining an image such as a human image. Here, the method of placing an evaluation weight on an image having a large degree of change may be a method of changing the weighting while counting, or a method of counting only pixels having a large degree of change.

The weighting of the evaluation criterion does not necessarily have to be fixed, and the invention according to claim 5 is a method according to any one of claims 1 to 4, wherein the weighting of each evaluation criterion is performed in the image evaluation method according to any one of claims 1 to 4. Is configured to be changeable.

In the invention according to claim 5 configured as described above, it is possible to derive an overall evaluation result corresponding to an image by changing the weight for each evaluation criterion. In this case, various modes are included, such as changing each weight individually or preparing a plurality of combinations in advance and selecting the combination.

Further, the change of the weighting itself is not performed by the operator but is realized based on the image data. As an example, the invention according to claim 6 is based on each evaluation criterion. It is configured to change the weight of the evaluation result based on the evaluation result.

In the invention according to claim 6, the evaluation result is obtained by each evaluation criterion, and the weight of the evaluation result is changed from the evaluation result in consideration of the adaptability of each evaluation criterion. Let it.

Various methods can be adopted when changing the weight of the evaluation criterion using the evaluation result. For example, it is determined whether or not the image data of each pixel is subjected to the above-described sampling with a certain evaluation criterion. If it is determined, the number of pixels is used as one evaluation criterion, and when the number of pixels is large, weighting is increased.

The idea of the invention for evaluating an image by the above-described method includes various aspects. That is, it can be appropriately changed, for example, by hardware or software.

In the case of software for performing image processing as an embodiment of the idea of the present invention, the software naturally exists on a software recording medium on which such software is recorded, and must be used.

As one example, the invention according to claim 7 is a computer which inputs image data of a real photograph composed of pixels in a dot matrix form by a computer, totals the image data of each pixel on a predetermined basis, and based on the totaling result. A medium on which an image evaluation program for evaluating an image is recorded, which has a plurality of evaluation criteria for the above-mentioned total result, and is configured to sum the evaluation results based on the respective evaluation criteria with a predetermined weight.

Of course, the recording medium may be a magnetic recording medium, a magneto-optical recording medium, or any software recording medium to be developed in the future. Also, the duplication stages of the primary duplicated product, the secondary duplicated product, and the like are equivalent without any question. In addition, the present invention is still used even when the supply method is performed using a communication line, and the same applies to a case where the information is written on a semiconductor chip.

Further, even when a part is realized by software and a part is realized by hardware, there is no difference in the concept of the invention, and it is necessary to store a part on a software recording medium. May be in a form that is read as appropriate in accordance with.

It goes without saying that the image evaluation method and the software can be applied as an image evaluation apparatus as a subject of implementation, and the invention according to claim 8 inputs actual image data composed of pixels in a dot matrix. Image data input means, and sums the image data of each pixel according to a predetermined standard, and evaluates an image based on the totalized result. And image data evaluation means for summing the data with a predetermined weight.

In the invention according to claim 8 configured as described above, the image data input means inputs real image data consisting of pixels in a dot matrix form, and the image data evaluation means converts the image data of each pixel to a predetermined value. And the image is evaluated based on the result of the aggregation. At this time, the image data evaluation means has a plurality of evaluation criteria for the result of the aggregation, and evaluates the evaluation results based on the respective evaluation criteria by adding a predetermined weight to the evaluation results.

Needless to say, such an image evaluation apparatus may be used alone or may be used in a state of being incorporated in an image processing apparatus.

[0033]

As described above, the present invention provides an image evaluation method capable of flexibly responding to the determination of image characteristics because the evaluation is comprehensively performed by changing the weights of a plurality of evaluation criteria. can do.

According to the second aspect of the present invention, since the image data is sampled and processed, a plurality of evaluation criteria can be adopted according to the sampling method, and the processing amount can be reduced. .

Further, according to the third aspect of the present invention,
It is possible to employ an evaluation criterion that is optimal for a landscape or the like while reducing the processing amount.

Further, according to the invention of claim 4,
Since a portion where the degree of change of the image is large is often a subject portion with clear focus, it is possible to perform image evaluation with emphasis on such important pixels.

Further, according to the invention of claim 5,
By changing the weights for a plurality of evaluation criteria, more flexible evaluation becomes possible.

Further, according to the invention of claim 6,
Since the weighting is changed using the evaluation result, it is possible to reduce the trouble of the evaluation.

Further, according to the invention of claim 7,
In the same manner, it is possible to provide a medium on which an image evaluation program capable of flexibly coping with determining the characteristics of an image has been recorded.
An image evaluation device can be provided.

[0040]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an image processing system for performing image processing by executing an image evaluation method according to an embodiment of the present invention, and FIG. 2 is a schematic block diagram showing an example of a specific hardware configuration. Is shown.

In FIG. 1, an image input device 10 outputs real image data representing a photograph or the like as pixels in a dot matrix to an image processing device 20 and outputs the data to the image processing device 2.
In step 0, the image data is totaled through a predetermined process to obtain an evaluation result, and the content and degree of the image processing are determined based on the evaluation result, and then the image processing is executed. The image processing device 20 outputs the image-processed image data to the image output device 30,
The image output device outputs the image-processed image as pixels in a dot matrix.

The image processing apparatus 20 collects image data in advance and obtains an evaluation result for the image. At this time, the image data is totaled individually by using a plurality of evaluation criteria, and the weights are changed under predetermined conditions to add up. Therefore, the image processing device 20 forms an image data evaluation unit.

A specific example of the image input device 10 corresponds to the scanner 11, the digital still camera 12, or the video camera 14 in FIG. 2, and a specific example of the image processing device 20 is a computer 21, a hard disk 22, a keyboard 23, a CD-ROM. A computer system including a drive 24, a floppy disk drive 25, a modem 26, and the like corresponds to the computer system. Specific examples of the image output device 30 correspond to a printer 31, a display 32, and the like. In the case of the present embodiment, actual image data such as a photograph is preferable as the image data in order to evaluate the image in order to correct a defect or the like of the image. The modem 26 is connected to a public communication line, is connected to an external network via the public communication line, and can download and install software and data.

In this embodiment, the image input device 10
The scanner 11 and the digital still camera 12 output RGB (green, blue, red) gradation data as image data, and the printer 3 as the image output device 30.
1 requires CMY (cyan, magenta, yellow) or CMYK binary data obtained by adding black to the input as gradation data, and the display 32 has RGB data.
Is required as an input. On the other hand, an operating system 21a is running in the computer 21, and a printer driver 21b and a display driver 21c corresponding to the printer 31 and the display 32 are provided.
Is incorporated. The image processing application 21d is controlled by the operating system 21a to execute processing, and executes predetermined image processing in cooperation with the printer driver 21b and the display driver 21c as necessary.

Therefore, the specific role of the computer 21 as the image processing device 20 is to execute RGB image processing while performing optimum image processing while evaluating the image by inputting RGB gradation data.
B, and generates the gradation data of the display driver 21c.
Via the printer driver 21b, and CMY (or CMY) via the printer driver 21b.
The data is converted into the binary data K) and printed by the printer 31.

As described above, in this embodiment, a computer system is incorporated between image input / output devices to perform image evaluation and image processing. However, such a computer system is not necessarily required. The present invention is applicable to a system that performs various types of image processing on image data. For example, as shown in FIG. 3, a digital still camera 12a incorporates an image processing device that performs image evaluation and image processing, and uses the converted image data to display on a display 32a or print on a printer 31a. May be. As shown in FIG. 4, in a printer 31b that inputs and prints image data without passing through a computer system, image evaluation is performed based on image data input through a scanner 11b, a digital still camera 12b, a modem 26b, or the like. It is also possible to configure so as to perform image processing.

The above-described image evaluation and the accompanying image processing are performed by the image processing program corresponding to the flowchart shown in FIG. The flowchart shown in the figure corresponds to the first part of the image evaluation in the image processing program, and executes a process of totalizing the image data based on a plurality of evaluation criteria and obtaining a predetermined evaluation result.

Here, two evaluation criteria used in this embodiment will be described. What is common is that not all pixels are targeted, but pixels are thinned out according to a predetermined criterion and the luminance of sampled pixels is totaled. The difference is that one samples the pixels uniformly while the other samples the edge pixels. The result of summation of the luminance will be described later, but by changing the so-called sampling method in this way, the evaluation of the image can be changed. Sampling pixels evenly means counting the brightness of the pixels in the entire image,
Since the distribution of the luminance of the image data of the entire image is evaluated, it can be used as a reference for the evaluation that the landscape photograph is generally dark or the contrast is narrow.

On the other hand, since the edge pixels are sharp portions of the image, the luminance of the pixels related to the original subject in the image is totaled. Even if the background is dark, the subject has sufficient brightness. Then, an evaluation result that the brightness of the image is sufficient can be obtained. In the present embodiment, an image is determined by appropriately combining these two evaluation criteria by selection by an operator or automatic processing.

Referring to the flowchart shown in FIG.
In this image evaluation process, as shown in FIG. 6, the target pixel is moved by performing sub-scanning in the vertical direction while performing main scanning in the horizontal direction with respect to image data composed of pixels in a dot matrix.
Each pixel is counted by judging whether or not it is a sampling target.

If the image data is composed of dot matrix pixels, the above-described RG
It is represented by gradation data representing the luminance of B, and the difference of the same data between adjacent pixels becomes large in the edge portion of the image. This difference is a luminance gradient, which is called an edge degree. In step S110, the edge degree at each pixel is determined. When considering the XY rectangular coordinates as shown in FIG. 7, the vector of the degree of change of the image is the X-axis direction component and the Y-axis direction component.
The calculation can be performed by obtaining the axial components.

In a digital image composed of pixels in the form of a dot matrix, as shown in FIG. 8, the pixels are adjacent to each other in the direction of the vertical axis and the horizontal axis, and the brightness is represented by f (x, y).
It shall be represented by In this case, f (x, y) is R (x, y), G (x, y), B (x,
y) or the whole luminance Y (x, y). Note that R (x,
y), G (x, y), B (x, y) and overall luminance Y
Strictly speaking, the relationship with (x, y) cannot be converted without referring to a color conversion table or the like, but a simple correspondence may be used as described later.

In FIG. 8, the difference value fx in the X direction and the difference value fy in the Y direction are

[0055]

(Equation 1)

Is represented as follows. Therefore, the magnitude | g (x, y) |

[0057]

(Equation 2)

Is represented as follows. Of course, the edge degree is represented by | g (x, y) |. Note that, as shown in FIG. 9, the pixels are originally arranged in a grid in the vertical and horizontal directions, and there are eight adjacent pixels when focusing on the central pixel. Therefore, similarly, the difference between the image data and each adjacent pixel may be represented by a vector, and the sum of the vectors may be determined as the degree of change of the image.

Since the edge degree is obtained for each pixel as described above, a pixel having a larger edge degree than a certain threshold value may be determined as an edge pixel. In addition,
Considering empirical facts, a subject with a high focus is often located at the center of the composition. Therefore, by adopting a structure in which a large number of pixels are extracted from the central portion, a more favorable effect can be obtained when used for determination of image processing.

For this reason, as shown in FIG. 10, the threshold values Th1, Th2, and Th3 to be compared for each portion in the image may be different. Of course, in this example,

[0061]

(Equation 3)

The threshold value is lower in the portion closer to the center, and it is determined that the focus is concentrated even if the edge degree is relatively low.

In step S120, it is determined whether or not the degree of change is large by comparing the edge degree with the same threshold value.
As a result of the comparison, if the degree of edge is larger, the pixel is determined to be an edge pixel, and in step S130, the image data of the pixel is sampled and stored in the work area. The work area may be the RAM in the computer 21 or the hard disk 22.

On the other hand, in parallel with the determination of the edge degree, it is determined in step S140 whether or not the target pixel is a target pixel for uniform sampling. Even if sampling is performed uniformly, it is necessary to perform sampling to such an extent that it is within a certain error range. According to the statistical error, the error with respect to the number N of samples can be approximately expressed as 1 / (N ** (1/2)). Here, ** indicates a power. Therefore, 1%
In order to perform processing with an error of the order, N = 10000.

Here, the bit map screen shown in FIG. 6 has the number of pixels of (width) × (height), and the sampling period ratio is

[0066]

(Equation 4)

Assume that Here, min (widt
h, height) is the smaller of width and height, and A is a constant. Also, the sampling period ratio here indicates how many pixels are sampled, and the pixels marked with a circle in FIG. 11 show the case where the sampling period ratio = 2. In other words, one pixel is sampled every two pixels in the vertical direction and the horizontal direction, and sampling is performed every other pixel. A =
The number of sampling pixels in one line when the number is set to 200 is as shown in FIG.

As can be seen from the figure, except for the case where the sampling period ratio = 1 in which sampling is not performed, when the width is 200 pixels or more, the number of samples is at least 100 pixels or more. Therefore, when the number of pixels is 200 or more in the vertical and horizontal directions, (100 pixels) × (100 pixels) = (10000 pixels)
Is ensured, and the error can be reduced to 1% or less.

Here, min (width, hei)
(ght) is based on the following reason.
For example, assuming that width >> height as in the bitmap image shown in FIG. 13A, when the sampling period ratio is determined by the longer width, as shown in FIG. 13B. In addition, it may happen that only two lines, the upper and lower lines, are extracted in the vertical direction. However, min (wi
(dth, height), if the sampling period ratio is determined based on the smaller one, it is possible to perform the thinning including the intermediate part even in the smaller vertical direction as shown in FIG. Become. That is, it is possible to perform sampling while securing a predetermined number of extractions.

In step S140, while employing such a uniform sampling method, it is determined whether or not the target pixel is to be sampled. If so, image data is sampled in step S150.

Sampling the image data in steps S130 and S150 means that the luminance is totalized based on the image data. As described above, in this embodiment, the computer 21 handles RGB
And does not directly have a luminance value. Although it is possible to perform color conversion to the Luv color space in order to obtain the luminance, this is not advisable due to problems such as the amount of calculation. For this reason, the following conversion formula for directly obtaining luminance from RGB used in the case of a television or the like is used.

[0072]

(Equation 5)

The luminance is tabulated as a histogram. Of course, the tabulation area of step S130 and the step S150
Are separate areas. It should be noted that the number of pixels to be counted is also counted together with the counting of the luminance.

In order to perform the above processing for each pixel of the image data, the pixel to be processed is moved in step S160, and the processing is repeated until it is determined in step S170 that all the pixels have been completed.

After summing the luminance of the target pixel by each sampling method, step S180
Now enter the image evaluation options. FIG. 14 shows an image evaluation option input screen displayed on the display 32, and three options of portrait, landscape photograph, and automatic setting are prepared as options.

As shown in FIG. 15, when generating a histogram for evaluating the image by adding the luminance histogram obtained by the uniform sampling and the luminance histogram obtained by sampling the edge pixels, Needs to be adjusted. Here, when the weighting coefficient k is adopted, the summation result D of the uniform sampling is obtained.
The evaluation result Dist_S is obtained from the ist_ave and the aggregation result Dist_edg in the edge pixel sampling.
um is

[0077]

(Equation 6)

Is obtained. And this weighting coefficient k is
It can be said that the closer to "0", the greater the importance of the entirety, and the closer to "1", the greater the importance of the subject. For this reason, after selecting an option on the image evaluation option input screen shown in FIG. 14, the process branches in step S190 based on the option, and when a portrait is selected, “k” is selected in step S192.
= 0.8 ", and when a landscape photograph is selected," k = 0.2 "is set in step S194.

One remaining option selection is automatic setting. In this automatic setting, as described above, based on the number of edge pixels sampled, if the number of edge pixels is small, the weighting coefficient k is set to be close to “0” in consideration of a landscape photograph, and if the number of edge pixels is large. , The weighting coefficient k is made closer to “1” as a portrait. Using the sampling number x_edg of the edge pixels and the uniform sampling number x_ave, the weighting coefficient is calculated in step S196.

[0080]

(Equation 7)

The total result Dist for calculation and evaluation
Get _Sum.

In this way, the evaluation result Dist
Obtaining _Sum means that the image has been evaluated. Of course, further determination may be made using this totaled result, but basically it may be changed as appropriate in accordance with image processing using this totaled result.

Thereafter, the optimum image processing is determined and executed based on the result of the aggregation. FIG. 16 shows a flowchart for executing image processing for contrast enhancement and brightness correction as an example.

In the basic method for increasing the contrast in the present embodiment, a luminance distribution is obtained based on image data, and this luminance distribution uses only a part of the original gradation width (255 gradations). If so, the distribution is expanded.

Therefore, in step S310, a histogram of the luminance distribution as the aggregation result Dist_Sum is created from the above-mentioned weighting coefficient k, and in step S320, the width to be enlarged is determined. In determining the enlargement width, consider obtaining both ends of the luminance distribution. The luminance distribution of the photographic image appears substantially in a mountain shape as shown in FIG. Of course, the position and shape are various. The width of the luminance distribution is determined depending on where the both ends are determined. However, the point where the number of distributions becomes “0” by extending the base cannot be set as the both ends. This is because the number of distributions may change around “0” in the tail part, and the number of distributions changes while approaching “0” without limit statistically.

For this reason, the portions extending inward by a certain distribution ratio from the side with the highest luminance and the side with the lowest luminance in the distribution range are set as both ends of the distribution. In the present embodiment, as shown in the figure, the distribution ratio is set to 0.5%. Of course, this ratio can be changed as appropriate. In this manner, by cutting the upper and lower ends by a certain distribution ratio, white points and black points caused by noise or the like can be ignored. That is, if such processing is not performed, even if there is even one point, a white point or a black point will be at both ends of the luminance distribution. "0"
And the uppermost end is the gradation “255”,
Such a problem is eliminated by setting a portion which is 0.5% of the number of pixels inside from the upper end portion as an end portion.

In actual processing, 0.5% of the number of sampled pixels is calculated, and the number of distributions is accumulated in order from the upper-end luminance value and the lower-end luminance value of the reproducible luminance distribution in order inward. The luminance value having a value of 0.5% is obtained. Hereinafter, this upper end is called ymax, and the lower end is called ymin.

The reproducible luminance range is from “0” to “25”.
When “5” is set, the luminance Y to be converted is obtained from the luminance y before the conversion and the maximum value ymax and the minimum value ymin of the luminance distribution range based on the following equation.

[0089]

(Equation 8)

However,

[0091]

(Equation 9)

If Y <0 in the above conversion formula, Y = 0.
If Y> 255, Y = 255. Here, “a” is a slope, and “b” is an offset. According to this conversion formula, as shown in FIG. 18, a luminance distribution having a certain narrow width can be expanded to a reproducible range. However, when the luminance distribution is expanded by making the most of the reproducible range, a highlight portion may be lost in white, or a high shadow portion may be lost in black. In order to prevent this, in this embodiment,
Reproducible range is limited. That is, the luminance value “5” is left as a range that does not expand to the upper and lower ends of the reproducible range. As a result, the parameters of the conversion equation are as follows.

[0093]

(Equation 10)

In this case, y <ymin and y>
No conversion is performed in the range of ymax.

However, if the enlargement ratio (corresponding to a) is applied as it is, a very large enlargement ratio may be obtained. For example, in a twilight state such as in the evening, the width of the contrast from the brightest part to the dark part is naturally narrow, but as a result of trying to greatly increase the contrast of this image, it is converted like a daytime image. I could end up. Since such conversion is not desired, a limit is set for the enlargement ratio, and a is set to 1.5 (to
2) Restrict it so that it does not exceed the above. Thereby, twilight comes to be expressed as twilight. In this case, processing is performed so that the center position of the luminance distribution does not change as much as possible.

By the way, it is irrational to execute the above conversion formula (Y = ay + b) every time the luminance is converted.
That is, the possible range of the luminance y is “0” to “25”.
Since it can be only 5 ", the converted luminance Y can be obtained in advance for all possible values of the luminance y. Therefore, it is stored as a table as shown in FIG.

Forming such a conversion table corresponds to the enlargement width determination processing in step S320, and it is possible to change the image data. However, since it is extremely effective not only to enhance the contrast by expanding the range of brightness but also to adjust the brightness together, the brightness of the image is determined in step S330.
Generate parameters for correction.

For example, in the case where the peaks of the luminance distribution as a whole are closer to the dark side as shown by the solid line in FIG. 20, it is better to move the peaks to the brighter side as a whole as shown by the wavy line. Conversely, when the peaks of the luminance distribution are generally shifted to the bright side as shown by the solid line in FIG. 21, it is preferable to move the peaks to the dark side as a whole as indicated by the wavy line.

As a result of various experiments, in the present embodiment, the median ymed in the luminance distribution is obtained, and when the median ymed is less than “85”, the image is determined to be a dark image and corresponds to the following γ value. Brighten with gamma correction.

[0100]

[Equation 11]

Alternatively,

[0102]

(Equation 12)

It is assumed that

In this case, even if γ <0.7, γ =
0.7. Unless such a limit is set, a night image looks like daytime. If the image is made too bright, the image becomes whitish as a whole and the image tends to have a low contrast. Therefore, a process such as emphasizing with saturation is preferable.

On the other hand, when the median ymed is larger than “128”, the image is determined to be a bright image and darkened by γ correction corresponding to the following γ value.

[0106]

(Equation 13)

Alternatively,

[0108]

[Equation 14]

It is assumed that In this case, even if γ> 1.3, a limit is set to γ = 1.3 so as not to be too dark.

Note that this γ correction may be performed on the luminance distribution before conversion or on the luminance distribution after conversion. FIG. 22 shows a correspondence relationship in the case of performing γ correction. If γ <1, the curve expands upward, and if γ> 1, the curve expands downward. Of course, the result of the γ correction may be reflected in the table shown in FIG. 19, and the correction is performed on the table data.

Finally, in step S340, it is determined whether contrast correction and brightness correction are necessary. In this determination, the enlargement ratio (a) and the γ value are compared with appropriate thresholds, and if the enlargement ratio is larger or the γ value exceeds a predetermined range, it is determined that the necessity exists. If it is determined that there is a need, the image data is converted.

When it is determined that image processing is necessary,
Conversion based on equation (9) is performed, and the conversion equation of the same equation is RG
This can also be applied to the correspondence relationship with the component values of B. The component values (R, G, B) after the conversion are compared with the component values (R0, G0, B0) before the conversion.

[0113]

(Equation 15)

Can be obtained as Here, the RGB component values (R0, G0, B0), (R, G, B) corresponding to the luminance y, Y of the gradation "0" to the gradation "255".
Are in the same range, and it can be said that the conversion table of the luminances y and Y described above may be used as it is.

Therefore, in step S350, the image data (R0, G0, B0) of all pixels are (18) to (2).
The process of referring to the conversion table corresponding to equation (0) and obtaining the converted image data (R, G, B) is repeated.

In this case, the result of summation of luminance is used as an evaluation criterion used for image determination, and contrast correction and brightness correction are performed. However, specific examples of image processing are not limited to this. Therefore, there are various kinds of aggregation contents used as evaluation criteria.

FIG. 23 is a flowchart showing a case where image processing for enhancing saturation is executed.

First, if the pixel data has saturation as its component element, it is possible to obtain the distribution using the saturation value. However, since the pixel data has only RGB component values, it is inherently necessary. Can not obtain a saturation value without conversion to a color space in which the saturation value is a direct component value. For example, in a Luv space as a standard color system, the L axis represents luminance (brightness), and the U axis and the V axis represent hue. Here, in the U-axis and the V-axis, since the distance from the intersection of both axes represents the saturation, substantially (U **
2 + V ** 2) ** (1/2) is the saturation.

In color conversion between such different color spaces, it is necessary to use an interpolation operation while referring to a color conversion table in which correspondence is stored, and the amount of operation processing becomes enormous. In view of such a situation, in the present embodiment, the alternative value X of the saturation is obtained as follows by directly using standard RGB gradation data as image data.

[0120]

(Equation 16)

Originally, the saturation becomes “0” when R = G = B, and reaches the maximum value when mixing a single color of RGB or a predetermined ratio of any two colors. Although it is possible to directly express the saturation directly from this property, the saturation of the maximum value can be obtained by a simple equation (21) if the color is a single color of red and yellow which is a mixed color of green and blue. It is "0" when the components are uniform. In addition, green and blue single colors also reach about half of the maximum value. Of course,

[0122]

[Equation 17]

The formula can be substituted.

In step S410, the distribution of the histogram for the alternative value X of the saturation is obtained separately while employing the above-described methods of uniform sampling and edge pixel sampling. In the equation (21), the saturation is distributed in a range from the minimum value “0” to the maximum value “511”, and the distribution is roughly as shown in FIG. Next step S42
At 0, a saturation index for this image is determined based on the aggregated saturation distribution. However, also in this case, the saturation index is individually derived from the totalization result of the uniform sampling and the totalization result of the edge pixel sampling, and the total saturation index is calculated using the above-mentioned weighting coefficient k.

In deriving the saturation index, in the present embodiment, the range occupied by the higher-order “16%” as the number of distributions in the range of the number of sampled pixels is obtained. Then, assuming that the lowest saturation “A” within this range represents the saturation of this image, the saturation enhancement index S is determined based on the following equation.

That is,

[0127]

(Equation 18)

It is assumed that FIG. 25 shows the relationship between the saturation “A” and the saturation enhancement index S. As shown in the drawing, the saturation index S is large when the saturation “A” is small and is small when the saturation “A” is large in the range of the maximum value “50” to the minimum value “0”. It will change gradually.

To enhance the saturation based on the saturation enhancement index S, if the image data has a saturation parameter as described above, the parameter may be converted. When the space is adopted, the space is temporarily converted to the Luv space which is a standard color system, and the Lu space is used.
It can be said that the displacement must be made in the radial direction in the v space. However, once the RGB image data is
After converting to v-space image data and emphasizing saturation,
Work such as returning to B is required, and the amount of calculation must be increased. Therefore, the saturation is enhanced using the RGB gradation data as it is.

When each component is a component value of a hue component having a substantially equal relationship as in the RGB color space, R = G
If = B, it is gray and achromatic. Therefore, R
Assuming that the minimum component of each component of GB simply reduces the saturation without affecting the hue of each pixel, the minimum value of each component is subtracted from all component values. It can be said that the saturation can be enhanced by enlarging the difference value.

First, a saturation enhancement parameter Sratio advantageous for calculation is calculated from the saturation enhancement index S described above.

[0132]

[Equation 19]

Is obtained. In this case, the saturation enhancement index S
When = 0, the saturation enhancement parameter Sratio = 1, and saturation enhancement is not performed. Next, assuming that the component value of blue (B) in each component (R, G, B) of the RGB gradation data is the minimum value, conversion is performed as follows using the saturation enhancement parameter Sratio.

[0134]

(Equation 20)

As a result, it is not necessary to perform two rounds of color conversion, which makes one round trip between the RGB color space and the Luv space, so that the calculation time can be reduced. In this embodiment, a method of simply subtracting the minimum value component from the other component values for the achromatic component is employed, but another conversion formula is employed for subtracting the achromatic component. It does not matter. However, in the case where only the minimum value is subtracted as in the equations (29) to (31), there is an effect that the amount of calculation becomes easy because no multiplication / division is involved.

Even when the formulas (25) to (27) are adopted, good conversion is possible, but in this case, when the saturation is enhanced, the luminance tends to be improved and the whole image tends to be brighter. Therefore, the conversion is performed on the difference value obtained by subtracting the luminance equivalent value from each component value.

First, in order to obtain the luminance, the above-described Lu is used.
Since the amount of calculation is large if the color conversion is performed in the v space, the RGB used in the case of television etc.
The following conversion formula for directly obtaining the luminance from is used.

The luminance Y is

[0139]

(Equation 21)

On the other hand, saturation enhancement

[0141]

(Equation 22)

It is assumed that: The addition and subtraction values △ R, △ G, and 求 め る B are obtained by the following equation based on the difference value with the luminance. That is,

[0143]

(Equation 23)

As a result,

[0145]

(Equation 24)

The conversion is possible. The preservation of the brightness is apparent from the following equation.

[0147]

(Equation 25)

When the input is gray (R = G = B), the luminance Y = R = G = B.
G = △ B = 0, and there is no achromatic color.
By using the expressions (39) to (41), the luminance is stored,
Emphasizing saturation does not make the whole picture brighter.

After the saturation enhancement index Sratio is obtained as described above, it is compared with a predetermined threshold value in step S430 to determine whether the image requires saturation enhancement. Then, if necessary, in step S440, equations (39) to
Image data is converted for all pixels based on equation (41).

Therefore, in steps S410 and S420, the image data is evaluated based on a plurality of evaluation criteria, and each evaluation result is added with a predetermined weight and added up. The configuration and the software constitute the image data evaluation means.

Further, the present invention can be applied to the evaluation of the edge degree based on the image processing of the edge enhancement processing as another image evaluation standard. FIG. 26 shows a flowchart of this edge enhancement processing. The edge degree is calculated by the above-described method. In step S510, the edge degree is separately calculated by the uniform sampling method and the edge pixel sampling method while moving the target pixel. Then, by dividing the integrated edge degree by the number of pixels, an average value of the edge degrees based on each evaluation criterion is calculated. That is, the sharpness SL of this image is obtained by setting the number of pixels to E.
(I) If pix,

[0152]

(Equation 26)

The calculation can be performed as follows. In this case, it can be determined that an image having a lower SL value has a lower degree of sharpness (visually blurred), and an image having a higher SL value has a higher degree of sharpness (visually clear).

Next, in step S515, a weighting coefficient k is determined by inputting an image evaluation option or the like, and the edge degrees based on each sampling method are weighted and added to be added.

On the other hand, since the sharpness of the image is intuitive, the sharpness SL is obtained in the same manner for the experimentally obtained image data having the optimum sharpness, and the obtained value is used as the ideal sharpness SLopt. In addition to the setting, in step S520, the degree of edge enhancement E
enhance,

[0156]

[Equation 27]

Is obtained. Here, the coefficient ks changes based on the size of the image. As described above, when the image data includes height dots and width dots in the vertical and horizontal directions,

[0158]

[Equation 28]

This is obtained as follows. put it here,
min (height, width) indicates the smaller one of the height dot and the width dot, and A is a constant “768”. Of course, these are obtained from the experimental results, and it goes without saying that they can be appropriately changed. However, good results are basically obtained by increasing the degree of emphasis for a larger image.

After the edge enhancement degree Eenhance is obtained in this manner, it is compared with a predetermined threshold value in step S530 to determine whether edge enhancement is necessary. If it is determined that edge enhancement is necessary, step S540 is performed. An edge enhancement process is performed on all pixels.

In the edge enhancement processing, the luminance Y ′ after emphasis is compared with the luminance Y of each pixel before emphasis.

[0162]

(Equation 29)

Is calculated. Here, Yunsharp is obtained by performing unsharp mask processing on the image data of each pixel. Here, the unsharp mask processing will be described. FIG. 27 shows an unsharp mask 41 of 5 × 5 pixels as an example. The unsharp mask 41 uses the value of “100” at the center as the weight of the pixel Y (x, y) to be processed in the matrix image data, and weights the peripheral pixels corresponding to the numerical values in the cells of the same mask. It is used for multiplying and integrating. When using this unsharp mask 41,

[0164]

[Equation 30]

The integration is performed based on the following arithmetic expression. (48)
In the formula, “396” is the total value of the weighting coefficients, and in the case of unsharp masks of different sizes, it is the total value of each cell. Mij is a weight coefficient described in the unsharp mask cell, and Y
(X, y) is image data of each pixel. Note that ij is indicated by horizontal and vertical coordinate values with respect to the unsharp mask 41.

The meaning of the edge emphasis operation calculated based on the expression (47) is as follows. Yunsharp
Since (x, y) is obtained by adding the peripheral pixel to the target pixel with a lower weight, so-called "unsharp" image data is obtained.
What has been blunted in this way has the same meaning as that obtained by applying a so-called low-pass filter. Therefore,
“Y (x, y) −Yunsharp (x, y)” has the same meaning as that obtained by subtracting a low-frequency component from all original components and applying a high-pass filter. Then, by multiplying the high-frequency component passed through the high-pass filter by the edge enhancement Eenhance and adding it to "Y (x, y)", the high-frequency component is increased in proportion to the edge enhancement Eenhance. The result is that the edges are enhanced.

Considering a situation where edge enhancement is required, since the image is a so-called edge portion of the image, the calculation may be performed only when the difference between the image data and the adjacent pixels is large. In this way, it is not necessary to perform the unsharp mask calculation on the image data portion which is not the most edge portion, and the processing is drastically reduced.

The actual calculation is based on the luminance Y ′ after emphasis and the luminance Y before emphasis.

[0169]

(Equation 31)

By replacing with R′G′B ′ after conversion,
Is

[0171]

(Equation 32)

The operation can be performed as shown in FIG.

Accordingly, in this edge enhancement processing, in steps S510 and S515, while evaluating the edge degree of the image based on a plurality of evaluation criteria, each evaluation result is given a predetermined weight and summed. The image data evaluation means is constituted by a hardware configuration and software for executing these.

It is determined whether to perform image processing for each of the above-described contrast correction, lightness correction, saturation enhancement, and edge enhancement. However, it is not always necessary to determine whether to perform image processing. That is, the degree of emphasis is set for each, and image processing may be performed with the degree of emphasis set in this way.

Next, the operation of the present embodiment having the above configuration will be described.

Assume that a photographic image is read by the scanner 11 and printed by the printer 31. Then, first,
Operating system 21a on computer 21
Is running, the image processing application 21
d is started, and the scanner 11 starts reading a photograph. When the read image data is captured by the image processing application 21d via the operating system 21a, the processing target pixel is set to the initial position. Subsequently, in step S110, equations (1) to
The edge degree is determined based on equation (3), and step S12 is performed.
At 0, the threshold is compared with the same edge degree. If the edge degree is higher, it is determined that the pixel to be processed is an edge pixel, and the image data of the pixel is stored in the work area in step S130. Step S1
In 40, it is determined whether or not the pixel to be processed is a target of uniform sampling.
If 0, the image data of the pixel is stored in another work area.

The above process is repeated while moving the pixel to be processed in step S160 until it is determined in step S170 that all the pixels have been executed.

When the execution has been completed for all the pixels, image data sampled with different evaluation criteria is stored in each work area.
At 180, an option for image evaluation is input. If the operator can determine whether the image is a portrait or a landscape by looking at the image, he or she can select either one. If the operator cannot determine whether the image is a portrait or a landscape, or if he / she wants to automate everything, he or she selects automatic setting. When the portrait is selected, the weighting coefficient k becomes “0.8” and the aggregation result of the edge pixels is weighted. When the landscape photograph is selected, the weighting coefficient k is “0.2”. When the automatic setting is selected, a weighting coefficient k corresponding to the ratio of the edge pixels is set. However, in any case, a plurality of evaluation criteria are adopted using the weighting coefficient k, and flexible evaluation can be performed without being restricted to only one evaluation criteria.

In the present embodiment, the image data itself is stored in the work area. However, it is not always necessary to store the image data itself in the work area from the viewpoint of memory capacity and processing time. That is, since a histogram of the luminance distribution and the saturation alternative value distribution is finally created for the sampling target pixel, the information of the histogram may be accumulated in advance in steps S120 and S150.

When the contrast correction and the brightness correction are automatically executed, the weighting coefficients are used to execute step S12.
At 0, S150, and S310, a histogram of the luminance distribution is obtained. At step S320, parameters for the enlargement process are determined based on the equations (12) and (13).
In step S330, parameters for brightness correction are determined based on equations (14) to (17). Then, in step S340, these parameters are compared with a predetermined threshold value.
At 0, luminance conversion is performed based on the above parameters. In this case, a luminance conversion table shown in FIG. 19 is first created in order to reduce the amount of calculation, and the image data is converted based on equations (18) to (20).

Thereafter, the image data subjected to the image processing is displayed on the display 32 via the display driver 21c. If the image data is good, the image data is printed by the printer 31 via the printer driver 21b. That is, the printer driver 21b receives the RGB tone data of which the edge is emphasized, performs rasterization corresponding to the print head area of the printer 31 through a predetermined resolution conversion, and performs color conversion of the rasterized data from RGB to CMYK. After that, the CMYK gradation data is converted into binary data and output to the printer 31.

With the above processing, the image data of the photograph read through the scanner 11 is automatically subjected to optimal contrast correction and lightness correction, displayed on the display 32, and printed by the printer 31. . That is,
It is possible to more flexibly determine an image by employing a plurality of evaluation criteria, and to realize optimal image processing such as contrast correction and brightness correction based on the evaluation result.

On the other hand, in the case of not only the contrast correction and the lightness correction but also the chroma emphasis and the edge emphasis, the chroma and the edge are sampled and aggregated by a plurality of evaluation criteria, and the weighting coefficient is adjusted. Therefore, the image processing is executed through a flexible determination that is not limited to a single evaluation criterion.

As described above, the computer 21, which is the center of the image processing, samples the image data of the pixels according to the different evaluation criteria in steps S120 and S140, and based on the image evaluation option input in step S180. In steps S192 to S196, a weighting coefficient k is determined, and the determined weighting coefficient k is used to add up the aggregation results in step S310 to generate a luminance distribution histogram, thereby obtaining a total of a plurality of evaluation criteria. The image is evaluated based on the statistical totaling result, and the optimal image processing can be executed in steps S310 to S350.

[Brief description of the drawings]

FIG. 1 is a block diagram of an image processing system to which an image processing device according to an embodiment of the present invention is applied.

FIG. 2 is a block diagram of specific hardware of the image processing apparatus.

FIG. 3 is a schematic block diagram illustrating another application example of the image processing apparatus of the present invention.

FIG. 4 is a schematic block diagram illustrating another application example of the image processing apparatus of the present invention.

FIG. 5 is a flowchart illustrating an image evaluation processing portion in the image processing apparatus of the present invention.

FIG. 6 is a diagram illustrating the size of image data and a state in which a processing target pixel is moved.

FIG. 7 is an explanatory diagram in a case where the degree of change of an image is represented by each component value of rectangular coordinates.

FIG. 8 is an explanatory diagram in a case where the degree of change of an image is obtained by a difference value between adjacent pixels in a vertical axis direction and a horizontal axis direction.

FIG. 9 is an explanatory diagram in a case where the degree of change of an image is obtained between all adjacent pixels.

FIG. 10 is a diagram showing a region where a threshold value is changed.

FIG. 11 is a diagram showing a sampling cycle.

FIG. 12 is a diagram showing the number of sampling pixels.

FIG. 13 is a diagram illustrating a relationship between a conversion source image and pixels to be sampled.

FIG. 14 is a diagram showing an input screen for an image evaluation option.

FIG. 15 is a diagram showing a situation in which individual sampling results are added with different weights.

FIG. 16 is a flowchart showing the second half of the image evaluation process and the image processing part.

FIG. 17 is a diagram showing edge processing of a luminance distribution and an edge obtained by the edge processing.

FIG. 18 is a diagram showing an enlarged luminance distribution and a reproducible luminance range.

FIG. 19 is a diagram showing a conversion table for enlarging a luminance distribution.

FIG. 20 is a diagram illustrating a concept of increasing brightness by γ correction.

FIG. 21 is a diagram illustrating a concept of darkening by γ correction.

FIG. 22 is a diagram illustrating a correspondence relationship of luminance changed by γ correction.

FIG. 23 is a flowchart in the case where saturation is emphasized.

FIG. 24 is a schematic diagram of a tallying state of a saturation distribution.

FIG. 25 is a diagram showing the relationship between saturation A and saturation enhancement index S.

FIG. 26 is a flowchart for edge emphasis.

FIG. 27 is a diagram showing an unsharp mask of 5 × 5 pixels.

[Explanation of symbols]

 Reference Signs List 10 image input device 20 image processing device 21 computer 21a operating system 21b printer driver 21c display driver 21d image processing application 22 hard disk 23 keyboard 24 CD-ROM drive 25 floppy disk drive 26 modem 30 ... Image output device

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] December 3, 1997

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0006

[Correction method] Change

[Correction contents]

In view of such a problem, the present applicant has proposed in Japanese Patent Application No. 9-151413 an invention for judging an important portion in an image. In the present invention, it is considered that an original subject (object) should exist in a sharp portion of an image, and a pixel having a large degree of change is determined as an object by focusing on the degree of change of the image at each pixel. doing.

Claims (8)

[Claims] 1. An image evaluation method for inputting real image data composed of dot matrix pixels, summing up image data of each pixel based on a predetermined standard, and evaluating an image based on the summation result. Having multiple evaluation criteria for the aggregation results,
An image evaluation method characterized in that evaluation results based on respective evaluation criteria are summed with a predetermined weight. 2. The image evaluation method according to claim 1, wherein when the image data is thinned out based on a predetermined criterion and totaled, sampling is performed based on a plurality of evaluation criteria, and the totaling result is calculated. An image evaluation method characterized by adding together with a predetermined weight. 3. The image evaluation method according to claim 2, wherein one evaluation criterion is one in which sampling is uniformly performed and tabulated. 4. An image evaluation method according to any one of claims 1 to 3, wherein one evaluation criterion is to aggregate the weights of the pixels having a large degree of change from adjacent pixels in each pixel. An image evaluation method, characterized in that: 5. An image evaluation method according to claim 1, wherein the weighting for each evaluation criterion can be changed. 6. The image evaluation method according to claim 1, wherein a weight of the evaluation result is changed based on an evaluation result based on each evaluation criterion. . 7. An image evaluation program for inputting image data of a real photograph made up of dot matrix pixels by a computer, summing up image data of each pixel based on a predetermined standard, and evaluating an image based on the summation result. Media that has multiple evaluation criteria for the above aggregation results,
A medium on which an image evaluation program is recorded, wherein evaluation results based on each evaluation criterion are summed with a predetermined weight. 8. An image data input means for inputting real image data consisting of pixels in a dot matrix form, and summing up image data of each pixel based on a predetermined standard and evaluating an image based on the summation result. An image evaluation apparatus, comprising: a plurality of evaluation criteria for a totaled result; and image data evaluation means for summing evaluation results based on each evaluation criteria with a predetermined weight.