JP2000306088A - Image correcting device and recording medium storing image correcting program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suitably correct an image which is photographed under adverse conditions and an image which has deteriorated by correcting the gradation width of a histogram by comparing the rates of the numbers of pixels by the gradations of the histogram with a threshold based upon user's specification. SOLUTION: A CPU calculates the numbers of pixels by the gradations of image data from the image data and calculates the rates Fi of the numbers of pixels of the gradations to the total number of pixels by the gradations. Then the rates Fi of the number of pixels are compared with the threshold S. The threshold S is previously calculated according to the specification input of a threshold done previously by a user. When the rates Fi of the numbers of pixels of successive gradations are both less than the threshold S, the integral value of those rates Fi of the numbers of pixels is regarded as the rate of the number of pixels of the gradations at the time of the generation of a histogram in an intermediate stage of correction when exceeding the threshold S. Consequently, a correct image holds gradation data of an uncorrected image below the threshold S in proportion to the rate Fi of the number of pixels and the data are not completely lost, so that characteristics of the original image are taken over.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image correction apparatus for correcting gradation characteristics of an image, and more particularly, to an illumination condition for capturing an image with an electronic camera using a solid-state image sensor or a camera using a silver halide film, The present invention relates to an image correction device for an image deteriorated due to an exposure condition and the like, and a machine-readable recording medium storing an image correction program.

[0002]

2. Description of the Related Art Generally, an image picked up by a camera includes a gray scale (brightness) for each pixel for both a monochrome image and a color image.
Is evaluated by Here, the gradation is black and white gradation for a black and white image, R (red) and G for a color image.
(Green), B (blue), or the gradation of each component such as C (cyan), M (magenta), and Y (yellow) is applied.
These images may be deteriorated due to photographing conditions, fading, etc., and image correction is performed by various methods.

A prominent method of image correction is gradation correction, and conventionally, a method using a histogram showing a gradation frequency distribution has been adopted. This histogram is calculated as follows. The gradation of each pixel constituting the image is detected, and the number (frequency) of pixels showing each gradation of, for example, 256 gradations is calculated. A histogram is obtained by plotting the number of pixels (frequency) for each gradation.

In correcting an image, conventionally, from each of an upper end (maximum gradation) and a lower end (minimum gradation) of a histogram, a gradation corresponding to the number of pixels specified by the user as a threshold is deleted, and the following equation 1 is used. The tone distribution of the image is corrected by converting the histogram according to the above conversion formula. C (x) = n (x−a) / (ba) Equation 1 where x is the gradation before correction, a is the minimum value of the gradation before correction, and b is the maximum of the gradation before correction. The value, C (x), is the corrected gradation, and n is the difference between the maximum gradation and the minimum gradation of the corrected image, and is normally set to the maximum value of 8-bit data, 255.

What is performed by the correction using the conversion formula of Expression 1 is, as can be easily understood, a work of extending the width of the gradation of the deteriorated image to 8 bits (256 gradations).

[0006]

In such a conventional technique,
For an image with a reduced contrast that exhibits a histogram as shown in FIG.
, The image is sufficiently image-corrected. In other words, the conventional technique is effective when there is no large constriction in the histogram.

However, depending on the lighting conditions and exposure conditions to be photographed, the conventional image correction method may not be able to correct the image. For example, when photographing a portrait in backlight, the phenomenon that the portrait is crushed black and the background is white and color jumps often occurs. A histogram of such an image is shown, for example, in FIG. From FIG. 6, it can be seen that the number of pixels of the intermediate gradation is extremely reduced as compared with the low gradation and the high gradation. The fact that this image cannot be sufficiently corrected by the conventional formula 1 can be understood from the fact that the gradation width of this image is already almost 256 gradations. The problem that the human image is crushed black and the background is white and jumps is hardly improved.

Further, as shown in FIG. 7, it is difficult to perform a sufficient correction even when a gradation region having a small number of pixels is widely spread at the lower end (generally, the upper and lower ends) of the histogram. That is, in the method of deleting from the upper end (maximum gradation) and the lower end (minimum gradation) of the histogram to the gradation corresponding to the number of pixels specified by the user as the threshold value, when the setting of the ratio of the number of pixels is low, For example, when the ratio up to the gradation corresponding to (L1) in FIG. 7 is set, a sufficient image correction effect cannot be obtained. Also, when the setting of the pixel number ratio is high,
For example, when the ratio up to the gradation corresponding to (L2) in FIG. 7 is set, there is a problem that a large part of the gradation information of the original image is lost. As described above, in some cases, an appropriate threshold value cannot be specified by the conventional method.

An object of the present invention is to provide an image correction apparatus for correcting an image which cannot be corrected by the conventional image correction method as described above, and a machine-readable recording medium storing an image correction program. is there.

[0010]

Means for Solving the Problems To solve the above problems,
In the conventional image correction method, the threshold for cutting off the histogram is specified by the ratio of the cumulative number of pixels of each gradation from the top or bottom of the histogram to the total number of pixels. On the other hand, in the present invention, a threshold is designated by the ratio of the number of pixels per gradation to the total number of pixels, and the image is corrected based on a comparison between the threshold and the gradation of the histogram.

Therefore, an image correction device according to the present invention comprises an image acquisition unit for acquiring image data from outside, and a histogram creation unit for calculating a ratio of the number of pixels to the total number of pixels for each gradation from the image data to create a histogram. A gradation correction unit that corrects a gradation width of a histogram based on a comparison between a threshold value of a pixel number ratio per gradation determined based on a user's designation and a pixel number ratio of each gradation in the histogram. By achieving the above, the above object is achieved.

The gradation correcting section can correct the gradation width of the histogram by compressing the width of the gradation in the histogram in which the ratio of the number of pixels is less than the threshold value. Here, the gradation width can be compressed by deleting the gradation whose pixel number ratio is less than the threshold value. The gradation width can be compressed at a compression ratio inversely proportional to the pixel number ratio. The width of the gray scale in which the ratio of the number of pixels is less than the threshold may be compressed to 1, which is the minimum gray scale width. That is, when there are a plurality of consecutive gray scales whose pixel number ratios are less than the threshold value, the integrated value of these pixel numbers may be combined into a frequency of one gray scale.

[0013] The threshold value is a function of gradation (i) (S (i)).
It may be. In this case, the ratio of the number of images of each gradation of the histogram is set to the threshold value S (i). This is the threshold (S)
Is changed in accordance with the gradation (i). For example, in the case of an image where importance is placed on the intermediate gradation, the threshold (S) may be decreased for the central portion, and the threshold (S) may be increased for the low gradation and the high gradation.

Further, the image processing apparatus may have an all gradation expanding section for expanding the gradation width of the entire histogram. A machine-readable recording medium on which the image correction program according to the present invention is recorded has an image acquisition procedure for acquiring image data from the outside, and a histogram is created by calculating the ratio of the number of pixels to the total number of pixels for each gradation from the image data. And a gradation for correcting the gradation width of the histogram based on a comparison between a histogram creation procedure to be performed and a threshold of the number of pixels per gradation determined based on a user's specification and a ratio of the number of pixels for each gradation of the histogram. A correction procedure program is stored. Thus, if the recording medium is set in the computer and the program is loaded, the computer can function as the image correction device.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 3 shows an overall configuration of an image correction apparatus using a computer according to an embodiment of the present invention. In FIG. 3, a computer 1 includes a CPU (microprocessor) 2. The input device 3 including a keyboard and a mouse, a hard disk 6, a memory 7, an image processing board 8, and an interface board 10 are connected to the CPU 2. A display device 9 is connected to an image output terminal of the image processing board 8. The interface board 10 includes an external device 11 such as an external recording medium.
Is connected. A CD-ROM drive device 12 is connected to the CPU 2, and a CD-ROM 13 in which an image correction program and an installation program thereof are recorded is inserted into the CD-ROM drive device 12.

According to the installation program in the CD-ROM 13, the CPU 2 develops the image correction program in the CD-ROM 13 and stores it in the hard disk 6 in an executable state. [First Embodiment] FIG. 1 is a flowchart for explaining the operation of an image correction apparatus according to a first embodiment of the present invention. Hereinafter, the operation of the image correction will be described with reference to the step numbers in FIG.

First, when the user starts the image correction program, the CPU 2 opens an image file to be corrected in step S1 to acquire image data, and proceeds to step S2. In step S2, the number of pixels is calculated for each gradation (i) from the image data. In the present embodiment, the gradation is 8-bit data, and the minimum value is 0 and the maximum value is 2
55.

In step S3, a ratio (Fi) of the number of pixels of each gradation (i) to the total number of pixels is calculated. FIG. 8 shows an example of a histogram of this calculation result.
In FIG. 8, the horizontal axis represents 256 levels of gradation (i), and the vertical axis represents the pixel number ratio (Fi). Here, the pixel number ratio (F
i) is represented by Equation 2. Pixel number ratio (Fi) = number of pixels of gradation (i) / total number of images Equation 2 In step S4, the variable TEMP in the temporary memory is reset to zero, and the process proceeds to step S5. Step S5
, The gradations i and j are set to i = 0 and j = 0. The gray scale i indicates the gray scale of the histogram of the original image, and the gray scale j indicates the gray scale of the histogram at the intermediate stage during the correction. In step S6, the flag FLAG is set to FAL which is the initial value.
Set to SE.

Thereafter, in step S7, the pixel number ratio Fi is compared with the calculated threshold value S. Threshold designation input performed by the user in advance (step S2
The threshold value S is calculated based on 3) (Step S)
24). As will be described later, for example, it is possible to specify whether the threshold is a constant regardless of the gradation or whether the threshold is a threshold corresponding to the gradation. If it is a constant, the input value is set as a threshold. On the other hand, when the threshold value is set according to the gradation, the threshold value S
Is defined as a gradation function S (i), and a threshold S
(I) is calculated.

In step S7, if the pixel number ratio Fi is less than the threshold value S, in step S8, the flag FLA
G is set to TRUE, and in step S9, the pixel number ratio Fi of the current gray scale i is added to the variable TEMP in the temporary memory. Then, in step S10,
The variable TEMP is compared with the threshold S. When step S10 is executed for the first time, the determination result is always NO.

If a negative determination is made in step S10, a variable TEMP is stored in variable Gj in step S10A, and the flow advances to step S11. In step S11, the gradation i is set to 1
In step S12, it is determined whether the gradation i is equal to or greater than 255. If step S12 is denied, the process returns to step S7. At this time, if a negative determination is made in step S7, steps S8, S9, S10, S1 are performed.
1. Repeat S12.

When step S10 is affirmed, in step S13, the value of the variable TEMP is stored in the variable Gj. That is, when the pixel number ratios Fi of a plurality of continuous gradations are both smaller than the threshold value S, the pixel number ratios Fi are integrated in step S9. This integral value is a threshold value S
Then, this integral value is set as the pixel number ratio of the gradation j at the time of creating the histogram at the intermediate stage during the correction.

When the process proceeds from step S13 to step S14, the variable TEMP is reset to zero, and the process proceeds to step S1.
At 5, one is added to the gradation j. Next, in step S22, the flag FLAG is set to FALSE. Thereafter, via steps S11 and S12, step S7
Return to The operation when step S7 is denied is as described above. If step S7 is affirmative, that is, if the pixel number ratio Fi with respect to the current gradation i is equal to or larger than the threshold value S, in step S16, the flag FLAG is set to TR.
It is determined whether or not it is a UE.

In step S16, the flag FLAG is set to F
It is determined whether it is ALSE. This step is provided to change the subsequent processing procedure depending on whether the pixel number ratio Fi of the previous gradation is less than the threshold value S or not less than the threshold value S. That is, if the pixel number ratio Fi of the previous gradation is equal to or larger than the threshold value S, a negative determination is made in step S16. In step S17, the pixel number ratio Fi of the current gradation i is stored in the variable Gj. , 1 to the gradation j. Then, steps S22, S
The operation in the case where a negative determination is made in step S16 via steps S11, S12, and S7 is as described above.

If the pixel ratio Fi of the previous gradation is less than the threshold value S, an affirmative determination is made in step S16, and step S16 is determined.
At 18, the value of the variable TEMP is stored in the variable Gj. Next, the process proceeds to step S19, where the value of the pixel number ratio Fi is stored in a variable Gj + 1. Further, step S20
Then, the variable TEMP is reset to zero, and step S2
1 adds 2 to the gradation j. Then, in a step S22, the flag FLAG is set to FALSE. The operation in the case where a negative determination is made in step S12 via step S11 has already been described.

An affirmative determination is made in step S12 for the acquired pixel ratio of the last gradation, and the process ends. Note that before step S10 is affirmed, step S10
When the process proceeds to A, S11, S12, S7, S16, S18, and S19, the integrated value of the pixel ratio Fi is set to the threshold value S before that.
Are not compressed into one gradation.

FIG. 9 shows a histogram of an image corrected by the operation of the flowchart of FIG. In FIG. 9, the horizontal axis represents 256 levels of gradation (j), and the vertical axis represents the pixel number ratio (Fi). That is, FIG. 9 is created based on the variable Gi obtained according to the flowchart of FIG. As can be seen from this histogram, by the above-described processing, the data of the gradation in which the ratio Fi of the number of pixels is less than the threshold value S is compressed. Accordingly, since the width between the maximum gradation and the minimum gradation is less than 256, this histogram is expanded to the maximum width 256 next. The conversion formula for the expansion is:
Is a known formula 3 similar to

C (x) = n (x−a) / (ba) Equation 3 where x is the gradation before conversion corrected by the operation in FIG. 1, and C (x) is the gradation after conversion. 1, a is the minimum value of the gradation corrected by the operation of FIG. 1, b is the histogram of the maximum value of the gradation corrected by the operation of FIG. 1, and n is the maximum gradation and the minimum gradation of the converted image. This is the maximum value of the 8-bit data, which is 255 here.

As described above, according to the method shown in the flowchart, a compression method for the gradation is established from the relationship between the gradations i and j. Furthermore, a method for extending the gradation to the maximum width is established by the above-mentioned known formula 3. When the correction processing combining the two methods of compression and expansion is performed on the histogram of FIG. 8, the histogram of FIG. 10 is obtained as a correction result, and it can be confirmed that the correction that effectively uses the gradation is performed. . In FIG. 10, the horizontal axis is 256
The gradation (i) of the stage and the vertical axis represent the pixel number ratio (Fi).
By performing a correction process combining the two methods of compression and expansion on the original image, corrected image data that effectively uses gradation is obtained, and this image data is stored as a new file.

In displaying an image, the CPU 2 opens the corrected image data file and displays it on the display device 9 via the image processing board 8. The image before the correction and the image after the correction are sometimes displayed side by side on the display device 9. Although the first embodiment of the present invention has been described above, in addition to the above-described method, the width of the gray scale in which the pixel number ratio is lower than the threshold is inversely proportional to the integral value of the pixel number ratio between the widths. Compressing at a compression ratio is also a preferable process. In other words, this method is a method of simply obtaining the width of the gradation in which the pixel ratio is lower than the threshold value, compressing the width, and compressing the width with a single compression ratio that is inversely proportional to the integral value of the ratio of the number of pixels therebetween.

In the conventional correction processing, the gradation data at the upper end and the lower end of the histogram of the image before correction are completely lost by the image correction. However, the image corrected in this embodiment is the same as the image before correction. There is a feature that tone data equal to or less than the threshold value is retained in proportion to the ratio of the number of pixels, is not completely lost, and tone correction can be performed while further inheriting the characteristics of the original image. [Second Embodiment] FIG. 2 is a flowchart for explaining the operation of the second embodiment of the image correction apparatus according to the present invention. Less than,
The operation of the image correction will be described with reference to the step numbers in FIG. In FIG. 2, steps similar to those in FIG. 1 are denoted by the same reference numerals. Embodiment 2 Including the portion overlapping with FIG.
Will be described.

First, when the user starts the image correction program, the CPU 2 opens an image file to be corrected in step S1 to acquire image data, and proceeds to step S2. In step S2, the number of pixels is calculated for each gradation (i) from the image data. In the present embodiment, the gradation is 8-bit data, and the minimum value is 0 and the maximum value is 2
55.

In step S3, a ratio (Fi) of the number of pixels of each gradation (i) is calculated. FIG. 8 shows an example of a histogram of this calculation result. Next, in step S4, the variable TEMP in the temporary memory is reset to zero, and the process proceeds to step S5. In step S5, the gradations i and j are set to i = 0 and j = 0, and in step S6
To set the flag FLAG to FALSE. Then, in step S7, the pixel number ratio Fi is compared with the calculated threshold value S. Based on the designation input of the threshold value (Step S23) performed by the user in advance,
The threshold value S has been calculated (step S24).

If the pixel number ratio Fi is smaller than the threshold value S in step S7, the flag FLA is set in step S8.
G is set to TRUE, and in step S9, the pixel number ratio Fi is added to the variable TEMP in the temporary memory. Then, the process proceeds to a step S9A, wherein the variable Gj is set to the variable T.
Substitute EMP and go to step S11. Step S1
In step 1, 1 is added to the gradation i, and in step S12, it is determined whether the gradation i is 255 or more. If step S12 is denied, the process returns to step S7.

If step S7 is affirmative, that is, if the pixel number ratio Fi is equal to or larger than the threshold value S, it is determined in step S16 whether the flag FLAG is TRUE.
If a negative determination is made in step S16, that is, the flag F
If LAG is determined to be FALSE, the value of the pixel number ratio Fi is stored in the variable Gj in step S17, and 1 is added to the gradation j in step S15. Thereafter, the process goes to step S16 via steps S11, S12, and S7.
Is negative as described above.

When the determination in step S16 is affirmative, that is, when the flag FLAG is TRUE, step S18
, The value of the variable TEMP is stored in the variable Gj. Next, the process proceeds to step S19, where the value of the pixel number ratio Fi is stored in a variable Gj + 1. Further, in step S20, the variable TEMP is reset to zero, and in step S21, 2 is added to the gradation j. Then, in step S22A, the flag FLAG is set to FALSE. Step S11
The operation in the case where a negative determination is made in step S12 via is described above.

An affirmative decision is made in step S12 for the last data obtained, and the process ends. FIG. 11 shows a histogram of an image corrected by the operation of the flowchart of FIG. In FIG. 11, the horizontal axis represents 256 gradations (j), and the vertical axis represents the pixel number ratio (Fi). As in the case of the first embodiment, in this histogram, since the data of the gradation whose pixel number ratio is less than the threshold value S is compressed, the width of the maximum gradation and the minimum gradation is less than 256.
Therefore, this histogram is expanded to a maximum width 256. The conversion equation for the enlargement is Equation 3.

When the correction processing combining the two methods of compression and expansion is performed on the histogram of FIG. 8, the histogram of FIG. 12 is obtained as a correction result, and the correction that effectively uses the gradation is performed. Is confirmed. By performing a correction process combining the two methods of compression and expansion on the original image, corrected image data that effectively uses gradation is obtained, and the image data is stored as a new file.

In displaying an image, the CPU 2 opens the corrected image data file and displays it on the display device 9 via the image processing board 8. The image before correction and the image after correction are sometimes displayed side by side on the display device 9. Compared with the first embodiment, the corrected image of the second embodiment compresses all the gradation widths before the threshold value is exceeded to the minimum gradation width of 1. However, pixels having a pixel ratio higher than the threshold value are more likely to be compressed. There is a feature that gradation can be used effectively.

In the first and second embodiments, as described above, the threshold value of the pixel number ratio may be specified as a constant value or as a function depending on the gradation (i). . When specifying as a constant value, the user can easily change the degree of image correction only by changing the magnitude of the constant. The larger the threshold value, the stronger the degree of image correction. In the case of designating as a function, it is a preferred embodiment to select a function form in which a portion with low brightness is small and a portion with high brightness is large. The reason is that an image with low brightness (low gradation) has a slight change in gradation more conspicuous than an image with high brightness (high gradation).

In the first and second embodiments,
The designation of the threshold value (step S23) may be performed manually by the user each time the process is performed, or may be performed by being preset and automatically called. As described above, the first embodiment
The present invention has been described with reference to the second embodiment. In this description, the histogram is compressed and then expanded to 256 gradations.
The present invention is not limited to this order. Compression and expansion may be performed together, or conversely, expansion may be performed first, and then compression may be performed later.

[0042]

As described above, according to the present invention, the threshold is specified by the ratio of the number of pixels per gradation to the total number of pixels.
Since the image is corrected based on the comparison between the threshold value and the gradation of the histogram, not only an image photographed under a bad condition under a special condition but also a deteriorated image can be preferably corrected.

[Brief description of the drawings]

FIG. 1 is a flowchart of a main part of Embodiment 1 of an image correction device according to the present invention.

FIG. 2 is a flowchart of a main part of a second embodiment of the image correction device according to the present invention.

FIG. 3 is a schematic diagram of a hardware configuration of the present invention.

FIG. 4 is an example of a histogram of an image.

FIG. 5 is a histogram of FIG. 4 corrected by a conventional method.

FIG. 6 is an example of a histogram that cannot be corrected by the conventional method.

FIG. 7 is another example of a histogram that cannot be corrected by the conventional method.

FIG. 8 is a histogram before correction of an image to be corrected (original image) used for describing the present invention.

FIG. 9 is an intermediate-stage histogram corrected by the image correction device according to the first embodiment of the present invention.

FIG. 10 is a histogram corrected by the image correction device according to the first embodiment of the present invention.

FIG. 11 is an intermediate-stage histogram corrected by the image correction device according to the second embodiment of the present invention.

FIG. 12 is a histogram corrected by the image correction device according to the second embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 computer 2 CPU 3 input device 7 memory 8 image processing board 9 display device 10 interface board 11 external device 12 CD-ROM drive device 13 CD-ROM 14 image before correction 15 image after correction

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification code FI Theme coat ゛ (Reference)

Claims (12)

[Claims] An image acquisition unit configured to acquire image data from outside; a histogram creation unit configured to calculate a ratio of the number of pixels to a total number of pixels for each gradation from the image data and create a histogram; A gradation correction unit that corrects the gradation width of the histogram based on a comparison between the determined threshold value of the number of pixels per gradation and the ratio of the number of pixels for each gradation of the histogram. Image correction device. 2. The image correction apparatus according to claim 1, wherein the gradation correction unit compresses a gradation width in which a ratio of the number of pixels in the histogram is less than the threshold. 3. The image correction apparatus according to claim 2, wherein the gradation correction unit compresses the gradation width at a compression ratio inversely proportional to the pixel number ratio. 4. The image correction apparatus according to claim 2, wherein the gradation correction unit compresses a gradation width in which the ratio of the number of pixels is less than the threshold value to a minimum gradation width of 1. . 5. The method according to claim 5, wherein the threshold value is a function of a gradation (i) (S
The image correction apparatus according to any one of claims 1 to 4, wherein (i)) is set, and a threshold value S (i) is compared with a pixel number ratio at each gradation of the histogram. 6. The image processing apparatus according to claim 1, further comprising an all gradation expanding section for expanding a gradation width of the entire histogram.
The image correction device according to any one of claims 1 to 5. 7. An image acquisition procedure for acquiring image data from outside, a histogram creation procedure for calculating a ratio of the number of pixels to the total number of pixels for each gradation from the image data and creating a histogram, and determining based on a user's designation. A gradation correction procedure for correcting the gradation width of the histogram based on the comparison between the threshold value of the pixel number ratio per gradation and the pixel number ratio for each gradation of the histogram. . 8. The recording method according to claim 7, wherein the gradation correction step compresses a gradation width in which a ratio of the number of pixels in the histogram is less than the threshold value. Medium. 9. The recording medium according to claim 8, wherein the gradation correction step compresses a gradation width at a compression ratio inversely proportional to the pixel number ratio. 10. The gradation correction step is such that the width of the gradation in which the ratio of the number of pixels is less than the threshold is a minimum gradation width of 1
9. A recording medium on which the image correction program according to claim 8 is compressed. 11. The method according to claim 1, wherein the threshold value is a function of a gradation (i) (S
11. The recording medium according to claim 7, wherein a threshold value S (i) and a pixel number ratio are compared for each gradation of the histogram. . 12. The recording medium on which the image correction program according to claim 7, further comprising a whole gradation expanding procedure for expanding a gradation width of the entire histogram.