JP2002281510A - Image processor and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an excessive exposure correction by performing the exposure correction by changing an exposure correction quantity corresponding to the rate of existence of an exposure aptitude part. SOLUTION: A dynamic range expanding means 102 corrects image data inputted from an image input device for expanding the R, G and B frequency distribution of the image data decomposed for the unit of a pixel into reproducing area. Corresponding to the rate of existence of a highlight part in the image, an exposure correcting means 103 changes the exposure correction quantity and performs the exposure correction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to an image processing apparatus and a program.

[0002]

2. Description of the Related Art JP-A-11-136514 discloses that
A technique is disclosed in which a portion having a large edge change is detected as a main subject from uniform sampling data, and an exposure correction amount is set from a median thereof. It also discloses that exposure correction is performed using a γ conversion curve.

Japanese Patent Application Laid-Open No. 2000-13624 has a plurality of highlight detection methods, in which highlight points are selected in accordance with the characteristics of an original image, and color processing conditions (exposure correction for changing image brightness) are disclosed. A technique for making the processing different is disclosed.

[0004]

Conventional cameras, digital cameras, and the like are provided with an automatic exposure control device for always maintaining the optimum exposure during photographing. In the exposure control method, the screen is divided into a plurality of appropriate areas for light amount detection,
A method of weighting each area, taking a weighted average, and adjusting the aperture, shutter speed, and the like is general. A method of simply taking a weighted average of the entire screen is called an average photometry method.
In addition, the main subject is often located at the center of the screen during photographing, so that the method of focusing on the central portion is called a center-weighted metering method.

In these methods, the exposure is determined based on the average value or the median value of the brightness of the photometry area. However, the “average or median brightness” is not always the proper exposure of the main subject. For example, there is a light source directly behind the main subject, and there is a large difference in luminance between the background and the main subject. In the case of “night shooting”, the main subject will be undershot even when shooting in the “night scene mode” which is a function of the digital camera.

[0006] Alternatively, if a flash is used even during nighttime shooting, the shutter speed is fixed, and if the flash light is weak, the main subject will be under.

That is, the current automatic exposure control system in a digital camera cannot cope with various photographing conditions.

[0008] In recent years, with the spread of personal computers, the Internet, and home printers, opportunities for handling photographic images with digital data have been increasing, and digital photographic image data (hereinafter, simply referred to as "image data") has been automatically created. Many techniques for correcting exposure have been proposed.

[0009] Here, the exposure correction means that an image having an unsuitable brightness for a scene is adjusted to a brightness suitable for the scene. For example, a subject that is dark overall due to insufficient exposure, a subject that is dark due to backlight, or a subject that is overexposed is darkened. In exposure correction by a camera, a method of changing an aperture and a shutter speed in order to adjust the amount of incident light entering a lens is generally used.
In the case of a printer or a display, this means processing for optimizing the brightness of an output signal with respect to the brightness of an input signal using an input / output conversion function or the like.

In the technique disclosed in Japanese Patent Application Laid-Open No. H11-136514, an area having a large edge change is detected as a main subject from image data, and an exposure correction amount is set based on the median value. However, since the edge portion (that is, the portion where the contrast change is large) does not always coincide with the main subject, there is a problem that the median value cannot be determined as the feature amount of the main subject.

FIG. 9 shows a luminance histogram of an image which does not require exposure correction, and FIG. 10 shows a luminance histogram of an image which does not require exposure correction. Both images were taken in the dark and the median of the histograms is very low. However, it can be seen that an image that does not require exposure correction includes a certain portion of highlight in the portion indicated by the arrow in the figure.

That is, if the subject extraction is uncertain, the subject extraction is not performed, and it is sufficient to check whether or not a highlight portion exceeding a certain predetermined ratio exists in a certain region. Will not be.

Further, since the processing for obtaining an edge portion disclosed in Japanese Patent Application Laid-Open No. H11-136514 is complicated, the processing speed is increased.

In the technique disclosed in Japanese Patent Application Laid-Open No. H11-136514, exposure correction is performed using a γ-conversion curve. There is also a problem that it is done.

An object of the present invention is to prevent an excessive exposure correction by performing an exposure correction by changing an exposure correction amount according to a ratio of an exposure suitable portion.

An object of the present invention is to easily and quickly determine the necessity of exposure correction without extracting a subject or the like.
The purpose is to obtain an appropriate correction amount.

An object of the present invention is to obtain a high-definition correction effect according to a scene.

An object of the present invention is to prevent overexposure correction and maintain the atmosphere of the original image.

An object of the present invention is to perform contrast enhancement suitable for a scene.

An object of the present invention is to perform exposure correction while emphasizing the contrast in the low and middle regions, and to create an appropriate tone curve.

An object of the present invention is to maintain gradation reproducibility at the output level of a contact.

An object of the present invention is to prevent excessive exposure correction and improve the speed.

An object of the present invention is to perform high-definition correction.

[0024]

According to a first aspect of the present invention, there is provided an image processing apparatus for correcting image data to an optimum exposure, wherein R, R,
Dynamic range expansion means for correcting image data in order to expand the G and B frequency distributions to the reproduction region, and exposure correction means for performing exposure correction by changing the amount of exposure correction in accordance with the proportion of highlights in the image An image processing apparatus comprising:

Therefore, by performing the exposure compensation by changing the exposure compensation amount in accordance with the proportion of the exposure suitable portion,
Excessive exposure compensation can be prevented.

According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the exposure correction means performs exposure using a level value corresponding to a frequency integration value of a predetermined ratio in a histogram representing brightness. It is characterized in that it is determined whether or not correction is necessary, and when exposure correction is necessary, the correction amount is determined from the median of the histogram.

Therefore, by using the level value corresponding to the predetermined ratio of the frequency integrated value in the histogram representing the brightness, it is determined whether or not there is a highlight portion having a predetermined ratio or more, thereby extracting a subject or the like. The necessity of exposure correction can be easily and quickly determined. Further, by determining the correction amount from the median value of the luminance histogram, an appropriate correction amount can be obtained.

The invention described in claim 3 is the first or second invention.
In the image processing apparatus described in the above, the exposure correction means,
The ratio of the highlight portion for changing the exposure correction amount may be varied depending on the scene.

Therefore, by changing the proportion of the highlight portion for changing the exposure correction amount in a scene such as a backlight image or a night person image, a high-definition correction effect corresponding to the scene can be obtained. it can.

According to a fourth aspect of the present invention, in the image processing apparatus according to any one of the first to third aspects, the exposure correction means includes a degree of influence of processing by the dynamic range expansion means and a highlight. The exposure compensation is performed by changing the exposure compensation amount in accordance with the ratio of the portions present.

Therefore, it is possible to prevent excessive exposure correction and maintain the atmosphere of the original image by examining whether or not it is necessary to perform further exposure correction on a pattern that has been greatly influenced by the dynamic range expansion processing. it can.

According to a fifth aspect of the present invention, in the image processing apparatus according to any one of the first to fourth aspects, the dynamic range expanding means changes the correction at least according to a backlight state. And

Therefore, the scene can be divided into at least a true backlight and others, and the reproduction range of the dynamic range is different for each of them, so that the contrast enhancement suitable for the scene can be performed.

According to a sixth aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, the exposure correction means performs tone curve correction as the exposure correction. The shape of the curve is characterized in that the contact position is made different according to the scene or exposure, as a primary straight line in a low region of the input level and a secondary parabola convex upward in a high region.

Therefore, when correcting the image information that is biased toward the dark level, the tone curve is enhanced by using a straight parabolic curve in the low area and an upwardly convex secondary parabola in the high area to enhance the contrast in the low and middle areas. While performing exposure correction. Also, by making the contact point between the straight line and the parabola different according to the scene and the exposure state, it is possible to create an appropriate tone curve.

According to a seventh aspect of the present invention, in the image processing apparatus according to the sixth aspect, the exposure correction means is provided in the image processing apparatus.
The two tone curves are tangentially connected at a contact point.

Therefore, by connecting the straight line and the parabola tangentially at the contact point, the tone reproducibility can be maintained at the output level of the contact point.

According to an eighth aspect of the present invention, in the image processing apparatus according to any one of the first to seventh aspects, the exposure correction means continuously connects the highlight portion on a two-dimensional position coordinate plane of the image. It is characterized by being an area.

Therefore, by performing the exposure compensation by changing the exposure compensation amount in accordance with the proportion of the exposure suitable portion,
Excessive exposure compensation can be prevented. Further, the speed can be improved by creating a correction table and performing color conversion processing instead of performing conversion for each pixel.

According to a ninth aspect of the present invention, in the image processing apparatus for correcting image data to an optimal exposure, the R, G, and B frequency distributions of the image data decomposed in pixel units are expanded to a reproduction area. A dynamic range expansion correction coefficient calculating means for calculating a correction coefficient of image data, and an exposure correction coefficient for calculating a correction coefficient for performing exposure correction by changing an exposure correction amount according to a ratio of a highlight portion in an image. An image processing apparatus comprising: a calculation unit; and a conversion unit that creates a correction table using the dynamic range expansion correction coefficient and the exposure correction coefficient and performs a color conversion process.

Accordingly, by determining the necessity of exposure correction using only the significant highlight portion, it is possible to perform high-definition correction.

According to a tenth aspect of the present invention, there is provided a computer readable program for causing a computer to execute image processing for correcting image data to an optimum exposure. B
To perform high-definition correction by determining the necessity of exposure correction using only significant highlights and dynamic range expansion processing that corrects image data to expand the frequency distribution to the reproduction area Is possible. This is a program that causes an exposure correction process to change the exposure correction amount according to the proportion of highlight portions present in an image to perform exposure correction, and causes a computer to execute the exposure correction process.

Therefore, by performing the exposure compensation by changing the exposure compensation amount in accordance with the proportion of the exposure suitable portion,
Excessive exposure compensation can be prevented.

According to an eleventh aspect of the present invention, in the program according to the tenth aspect, in the exposure correction processing, the exposure correction is performed by using a level value corresponding to a predetermined ratio of frequency integration values in a histogram representing brightness. It is characterized in that it is determined whether or not it is necessary, and when exposure correction is necessary, the correction amount is determined from the median of the histogram.

Therefore, by using the level value corresponding to the predetermined ratio of the frequency integrated value in the histogram representing the brightness, it is determined whether or not there is a highlight portion having a predetermined ratio or more, thereby extracting a subject. The necessity of exposure correction can be easily and quickly determined. Further, by determining the correction amount from the median value of the luminance histogram, an appropriate correction amount can be obtained.

According to a twelfth aspect of the present invention, in the program according to the tenth or eleventh aspect, in the exposure correction processing, a ratio of a highlight portion for changing the exposure correction amount is determined according to a scene. It is characterized by being different.

Therefore, by changing the proportion of the highlight portion for changing the exposure correction amount in a scene such as a backlight image or a night person image, a high-definition correction effect according to the scene can be obtained. it can.

The thirteenth aspect of the present invention relates to the tenth aspect.
13. The program according to any one of 12, wherein the exposure correction processing is performed by changing the exposure correction amount according to a degree of influence of the processing by the dynamic range expansion unit and a ratio of a highlight portion. Is performed.

Therefore, it is possible to prevent excessive exposure correction and maintain the atmosphere of the original image by examining whether or not it is necessary to perform further exposure correction on a pattern having a large degree of influence of the dynamic range expansion processing. it can.

The invention according to claim 14 is the invention according to claim 10
13. The program according to any one of 13, wherein the dynamic range expansion processing changes the correction at least according to a backlight state.

Therefore, the scene can be divided into at least true backlight and others, and the reproduction range of the dynamic range is different for each of them, so that the contrast enhancement suitable for the scene can be performed.

The invention according to claim 15 is the invention according to claim 10
14. In the program according to any one of the fourteenth aspects, in the exposure correction processing, a tone curve correction is performed as the exposure correction. It is characterized in that the contact position is made different according to the scene and the exposure as a secondary parabola that is convex.

Therefore, when correcting the image information which is biased toward the dark level, the tone curve is enhanced by using a straight parabolic curve in the low area and an upwardly convex secondary parabola in the high area to enhance the contrast in the low and middle areas. While performing exposure correction. Also, by making the contact point between the straight line and the parabola different according to the scene and the exposure state, it is possible to create an appropriate tone curve.

According to a sixteenth aspect of the present invention, in the program according to the fifteenth aspect, in the exposure correction processing, the two tone curves are tangentially connected at a contact point.

Therefore, by connecting the straight line and the parabola tangentially at the contact point, the tone reproducibility can be maintained at the output level of the contact point.

The invention according to claim 17 is the invention according to claim 10
In the program according to any one of the sixteenth aspects, the exposure correction process is characterized in that the highlight portion is a continuous area on a two-dimensional position coordinate plane of an image.

Therefore, by performing the exposure compensation by changing the exposure compensation amount in accordance with the proportion of the exposure suitable portion,
Excessive exposure compensation can be prevented. Further, the speed can be improved by creating a correction table and performing color conversion processing instead of performing conversion for each pixel.

According to an eighteenth aspect of the present invention, there is provided a computer readable program for causing a computer to execute image processing for correcting image data to an optimum exposure. B
Dynamic range expansion correction coefficient calculation processing for calculating a correction coefficient of image data in order to expand the frequency distribution to the reproduction area, and performing exposure correction by changing the exposure correction amount in accordance with the proportion of highlights in the image Causing the computer to execute an exposure correction coefficient calculation process of calculating the correction coefficient of the above, and a conversion process of creating a correction table using the dynamic range expansion correction coefficient and the exposure correction coefficient and performing a color conversion process. It is a featured program.

Accordingly, by determining the necessity of the exposure correction using only the significant highlight portion, it is possible to perform a high-definition correction.

[0060]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described.

FIG. 1 is a functional block diagram showing a schematic configuration of an image processing apparatus 101 according to an embodiment of the present invention, and FIG. 2 is a block diagram showing an example of a specific hardware configuration.

In FIG. 1, the image input device 100 outputs to the image processing device 101 actual image data representing a photograph or the like as pixels in a matrix. When performing a plurality of image processes, the image processing apparatus 101 performs each image process (the dynamic range expansion unit 102, the exposure correction unit 1
In step 03), an evaluation value is calculated based on the optimum evaluation means, and image processing is executed using the evaluation value. The image output device 104 outputs the image data on which the image processing has been performed by using pixels in a dot matrix.

A specific example of the image input device corresponds to the scanner 200 and the digital camera 201 or the video camera 202 shown in FIG. 2, and specific examples of the image processing device are a computer 210, a hard disk 203, and a keyboard 2.
04, a CD-ROM drive 205, a floppy (registered trademark) disk drive 206, a computer system including a modem 207, and the like.
04 is a printer 209 or a display 208.
Is applicable.

In the present embodiment, actual image data such as a photograph is suitable as image data. The computer 210 reads the program from a storage medium such as a CD-ROM, installs the program on the hard disk 203,
The software implements the functions of the present embodiment. The modem 207 may be connected to a public communication line, connected via an external network, download programs and data, install the programs and data on the hard disk 203, and realize the functions of the present embodiment using the software. .

On the other hand, an operating system 212 is running in the computer 210, and a printer driver 214 and a display driver 213 corresponding to a printer and a display are installed. The image processing application 211 is controlled by the operating system to execute processing, and executes predetermined image processing in cooperation with the printer driver 214 and the display driver 213 as necessary. Therefore, the role of this computer as an image processing device is to input RGB tone data, create RGB tone data that has been subjected to optimal image processing, display it on the display 208 via the display driver 213, and The data is converted into CMY or CMYK binary data via the printer driver 214 and printed by the printer 209.

The above-described acquisition of the image evaluation value and the accompanying image processing are performed specifically in the computer shown in FIG.
The processing is performed by an image processing program corresponding to the flowchart shown in FIG.

Hereinafter, such processing will be described assuming a scene other than backlight.

First, image data is inputted by the input means (300). Next, pixels are sampled in order to expand the RGB frequency distribution to an appropriate reproduction region (301), and R
The maximum value Max of the GB frequency distribution is calculated as a highlight point, and the minimum value Min is calculated as a shadow point (308), and a coefficient for expanding the dynamic range by the following equation:
α and β are obtained (309).

Α = 255 / (Max−Min) β = {(1−α) · (255 · Min)} / {255− (Max−Min)} (1) Here, α is limited, and α is set. If> 1.7, α≡1.7.

When sampling a pixel, the brightness is calculated by the following equation to create a brightness histogram (30).
2).

Brightness: Y = 0.299 · R + 0.587 · G + 0.114 · B (2)

A level L1 corresponding to a cumulative frequency of 95% is calculated from the luminance histogram (302). L1 is a value for calculating an evaluation value L2 for determining whether or not a predetermined ratio or more of a highlight portion exists in the image data after the dynamic range expansion unit 102.

Further, the cumulative frequency of the luminance histogram is 50%
Is calculated (302).
M1 is a value for calculating an evaluation value M2 for determining a correction amount when the highlight portion does not exist and exposure correction is required.

Here, the exposure correction means that an image having an unsuitable brightness for a scene is adjusted to a brightness suitable for the scene. For example, a subject that is dark overall due to insufficient exposure, a subject that is dark due to backlight, or a subject that is overexposed is darkened. In a display system such as a printer or a display, it means a process for optimizing the brightness of an output signal with respect to an input signal having an inappropriate brightness using an input / output conversion function, a look-up table, or the like.

Next, the luminance value Y _IN (i) (i = 1) at an arbitrary pixel
.., N, N: the total number of pixels), a dynamic range correction coefficient C1 (i) is obtained for each pixel i using α, β calculated by equation (1), and the input signal (R _IN (i), G _IN (i), B _IN (i)) are enlarged and converted into a reproduction area (310).

C1 (i) = (α · Y _IN (i) + β) / Y _IN (i) (3) (R1 (i), G1 (i), B1 (i)) = C1 (i)・ (R _IN (i), G _IN (i), B _IN (i))… (4)

Next, the exposure correction amount is determined. Exposure correction is performed as follows, and correction using the tone curve f (x) is performed using equations (7) and (8) (306).

In the present embodiment, the exposure correction amount is determined by calculating (m1, m2) using the values of L2 and M2. Also,
A method for deriving the tone curve f (x) will be described later.

First, M2 for the value L2 and M1 after the dynamic range expansion of L1 is calculated (303).

L2 = α · L1 + β (5) M2 = α · M1 + β (6)

Next, using L2, the bright portion 5 of the entire image
It is checked whether or not the region corresponding to% exceeds the threshold level 100, and it is checked whether or not a highlight portion exists (3).
04) If no highlight part exists, the exposure correction amount (m1, m2) is determined from M2 (305).

(1) When threshold level ≦ L2: No exposure correction is performed (N in 304).

(2) Threshold level> L2 and 0 ≦ M2 ≦ 5 or
In the case of 120 ≦ M2: The image is regarded as a night view or a normal light image, and no exposure correction is performed (N in 304).

(3) When threshold level> L2 and 5 <M2 ≦ 30: Exposure correction is performed (Y in 304). A tone curve f (x) is created with M1 = 160 and m2 = 192 (305). Tone curve correction (306).

(4) If threshold level> L2 and 30 <M2 <120, exposure correction is performed (Y in 304). M1 = 130, m2 = m1 +
Create tone curve f (x) with (-(59/90) × M2 + (236/3)) (3
05). Tone curve correction (306).

Here, m1 and m2 in (3) and (4) are contact points to be described later.
(m1, m2).

The tone curve correction is performed as follows. Brightness value of each pixel after dynamic range correction: Y1
From (i), an exposure correction coefficient: C2 (i) is calculated by the following equation.

C2 (i) = Y2 (i) / Y1 (i) = f (Y1 (i)) / Y1 (i) (7) Input signals (R1 (i), G1 (i), B1 (i) )) With the following formula,
Exposure correction is performed (307).

(R2 (i), G2 (i), B2 (i)) = C2 (i) · (R1 (i), G1 (i), B1 (i)) (8) where dynamic range Various methods have been proposed for the method of enlarging, that is, the method for determining Max and Min, which are reproduction regions, and the restriction of α, but are not limited thereto. Also, when determining a reproduction area or obtaining a luminance histogram, it is not always necessary to use information on all pixels,
Information obtained by thinning or the like may be used. further,
In this example, the evaluation values are calculated from the luminance histogram created before the dynamic range expansion unit 102, and the evaluation values are calculated by using the equations.
Although the evaluation values after the dynamic range expansion unit 102 are calculated by performing the conversions in (5) and (6), there is a method in which a luminance histogram is obtained after the dynamic range expansion unit 102 and the evaluation values are calculated. Various methods have been proposed for determining how much a highlight portion exists in the entire image, but the method is not limited to this. Various methods have been proposed for determining the exposure correction method and the exposure correction amount, but are not limited to these.

As a result, the scene becomes dark (bright).
However, by changing the exposure correction amount in accordance with the ratio of the exposure appropriate portions and performing the exposure correction, it is possible to prevent excessive exposure correction.

Further, by judging the ratio of the highlight portion using the level value corresponding to the predetermined ratio of the frequency integrated value in the histogram representing the brightness, the subject can be easily and quickly exposed without extracting the subject. The necessity of correction can be determined.

Further, by determining the correction amount from the median of the luminance histogram, an appropriate correction amount can be obtained.

In the above example, for example, in the case of a backlight image, the evaluation value L1 is set to a level corresponding to the cumulative frequency 80% of the luminance histogram, and the threshold level is set to 140. Here, the percentage or threshold value of the cumulative frequency is not limited to this. This makes it possible to obtain a high-definition correction effect according to the scene by making the ratio of the highlight portion for changing the exposure correction amount different between a scene such as a backlight image and a night person image.

If a technique of determining the backlight type of the light image and making the gradation processing different according to the backlight type as described later is used, extreme underexposure, a large dynamic range in nighttime shooting, and backlight scenes are used. Correction is expected. In such a case, if further exposure correction is performed, the atmosphere of the original image may be lost, so it is examined whether or not exposure correction is necessary.

For example, for a backlight image, α> 1.3 or more and a luminance frequency integrated value of 80%
Is not performed when the level corresponding to is equal to or larger than the threshold value 140.・ For night person images, α> 1.5 or more and brightness frequency integrated value 9
When the level corresponding to 5% is equal to or greater than the threshold value 100, no exposure correction is performed. And

Here, the magnitude of the influence of the dynamic enlargement processing is examined by the α value, but this is not limited. Thus, when the influence of the dynamic range expansion processing is large, it is possible to prevent excessive exposure correction and maintain the atmosphere of the original image by examining whether or not it is necessary to perform further exposure correction.

Further, if a technique of judging the backlight type of the light image and making the gradation processing different according to the backlight type is used,
It is possible to automatically and appropriately vary the reproduction range of the dynamic range according to the scene. That is,
The scene is divided into "true backlight", "backlight other than true backlight", and "other", and the highlight point Max and dark point Min of the reproduction area are set as follows.

In the case of true backlight: Max... Bipolarization branch point in the luminance histogram. Min ... Minimum value with a frequency of 0.3% or more of the total number of pixels in the luminance histogram.

In the case of backlight other than true backlight: Max ... R
Maximum value having a frequency of 0.03% or more of the total number of pixels in each GB histogram. Min: Minimum value with a frequency of 0.3% or more of the total number of pixels in the luminance histogram.

Others: Max: maximum value having a frequency of 0.03% or more of the total number of pixels in each of the RGB histograms. Mi
n: 0.03% of the total number of pixels in each RGB histogram
Minimum value with the above frequency.

Here, various methods have been proposed for the scene discrimination method and the dynamic range expansion method, but these are not limited to them. In addition, the scene may be determined by an operator in addition to the automatic determination. This makes it possible to divide the scene into at least true backlighting and others, and make the setting range of the dynamic range different for each, thereby performing contrast enhancement suitable for the scene.

The exposure correction means 103 is a tone curve correction represented by the equations (7) and (8). The tone curve f (x) (x ≧ 0) of the equation (7) is calculated by the following method. calculate. FIG. 4 shows f
(x) is shown. However, It is.

At the contact (x, y) = (m1, m2), l1 (x) and l2
(x) is connected, l1 (x) is a primary straight line, and l2 (x) is a secondary parabola.

L1 (x) = (m2 / m1) · x (0 ≦ x <m1) (9) L2 (x) = (− b ± (b ^ 2−4 · a · c) ^ (1 / 2)} / (2 · a) (m1 ≦ x <1)… (10) where a = tanθ / sqrt (2) b = 2 · tanθ · (x / sqrt (2)-P) + P… ( 11) c = tanθ ・ (x ^ 2 / sqrt (2) -2 ・ P ・ x)-P ・ x P = ((1-m1) ^ 2 + (1-m2) ^ 2) ^ (1/2 ) / 2, and θ is set so that l2 (x) ≦ 1. θ is (m
This is an angle formed by a straight line connecting (m1, m2) and (1, 1) and a straight line connecting (m1, m2) and the vertex of the parabola. The solution selects a value such that 0 <l2 (x) ≦ 1.

In order to have real roots, a, b, and c satisfy the following conditions.

B ^ 2−4 · a · c ≧ 0 (12)

Next, the point of contact (x, y) = (m1, m2) has a different value depending on the scene.

[1] Backlit Image: For example, the luminance histogram median M2 after dynamic range correction is
Lift amount by the function shown in: Δ (Δ =-(63/80) · M2 + (441
/ 4)), and m1 = 115, m2 = m1 + Δ. The low and middle level brightness is corrected while maintaining the contrast of the dark part. However, the brightness histogram for the exposure determination determines the backlight type of the light image, and the range from the minimum value Min to the bifurcation branch point described in the description of the technique for changing the gradation processing according to the backlight type. Perform using This makes it possible to determine the exposure in the area from which the backlight portion has been removed, and to obtain a high-definition exposure correction amount.

[2] Images other than backlight: 5 <M2 ≦ 30 ... m1 = 160, m2 = 192. Since the correction curve has a characteristic of raising the highlight side, the contrast of the dark part is enhanced and the lifting is small, so that even if the night scene image is erroneously recognized, there is little adverse effect. When 30 <M2 ≦ 1 Using the luminance histogram median M2 after the dynamic range correction, the lifting amount Δ (Δ = − (59/90) · M2 + (236/3)) is determined by the function shown in FIG. And m1 = 130 and m2 = m1 + Δ. The low and middle level brightness is corrected while maintaining the contrast of the dark part.

Here, there are various methods for calculating l1 (x), l2 (x) and (m1, m2), but the method is not limited to this. As a result, when correcting image information that is biased toward the dark level, the shape of the tone curve becomes a straight line in the low region: l1
(x), by using an upwardly convex secondary parabola: l2 (x) in the high region, exposure compensation can be performed while enhancing the contrast in the low and middle regions. Further, by changing the contact point (x, y) = (m1, m2) between the straight line and the parabola according to the scene or the exposure state, an appropriate tone curve can be created.

The above-mentioned exposure correction means 103 is given by the following equation.
Equation (7) is the tone curve correction represented by (8),
The tone curve f (x) (x ≧ 0) is calculated by the following method.

FIG. 7 shows f (x). However, It is.

At the contact (x, y) = (m1, m2), l1 (x) and l2
(x) is tangentially coupled, l1 (x) is a primary straight line, and l2 (x) is a secondary parabola expressed as follows.

L1 (x) = (m2 / m1) · x (0 ≦ x <m1) (13) L2 (x) = (− b ± (b ^ 2−4 · a · c) ^ (1 / 2)} / (2 · a) (m1 ≦ x <1)… (14) where a = tanθ · (sinξ ^ 2) b = 2 · tanθ · sinξ · (cosξ · x-m1-P) + P・ Cosξ… (15) c = tanθ ・ [cosξ ・ x ・ {cosξ ・ x-2 ・ (m1 + P)} + m1 ・ (m1 + 2 ・ P) -P ・ (m2 + sinξ ・ x)] Yes, ψ = atan {(1-m2) / (1-m1)} ξ = atan (m2 / m1) φ = ξ… (16) θ = atan {(1/2) tanφ} P = (1- m1) / 2 ・ cosψ

To have real roots, a, b, and c satisfy the following conditions.

B ^ 2−4 · a · c ≧ 0 (17)

Here, various methods can be considered for calculating l1 (x) and l2 (x), but the method is not limited to this. Thereby, since the straight line l1 (x) and the parabola l2 (x) are tangentially connected at the contact point, the tone reproducibility can be maintained at the output level of the contact point.

When a printer is assumed as the image display device, for example, processing as shown in FIG. 8 is performed to improve the image processing speed. That is, instead of directly correcting the image data, the dynamic range correction coefficient is actually:
C1 (j) (j = 0, 1, 2,..., 255) and an exposure correction coefficient: C2 (j) are obtained (804), a correction table is created (805), and the color conversion parameter (CMY ( k)) is updated (806).

With this method, an image of about 1.2 million pixels is averaged.
Processing can be performed with an overhead of 0.12 seconds.

When the image display device is used as a display, the image data is not directly corrected as in the case of the printer in order to improve the image processing speed, but the dynamic range correction coefficient is actually C1 (j)) (j = 0, 1, 2,..., 255) and an exposure correction coefficient C2 (j), a look-up table is created, and color conversion (RGB) is performed.

As a result, the scene becomes dark (bright).
However, excessive exposure correction can be prevented by changing the amount of exposure correction in accordance with the ratio of the exposure appropriate portions and performing the exposure correction. Also, the speed can be improved by creating a conversion table and performing color conversion processing instead of performing conversion for each pixel.

In the above example, the presence / absence of the highlight portion is determined only from the luminance histogram. However, if these portions are dispersed in the image, it cannot be said to be the main subject and the importance is low.
As shown in FIGS. 9 and 10, the highlight portion is significant only when it is concentrated on a certain image area and is established as a subject. Therefore, it is necessary to check whether the highlight portions are concentrated.

For example, as shown in FIG. 11, for example, as shown in FIG. 11, the image is divided into several areas and numbering is performed to determine whether the highlight area exists. Just check it. Here, various methods can be considered to check whether or not the highlight portion is concentrated in a certain area, but the method is not limited to this.

As a result, it is possible to determine the necessity of the exposure correction using only the significant highlight portion and to perform the high-definition correction.

[0125]

According to the first aspect of the present invention, excessive exposure correction can be prevented by changing the amount of exposure correction in accordance with the ratio of the exposure-appropriate portions and performing the exposure correction.

According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, a highlight value equal to or higher than a predetermined ratio is used by using a level value corresponding to a predetermined ratio of a frequency integrated value in a histogram representing brightness. By judging whether or not a part is present, it is possible to easily and quickly judge the necessity of exposure correction without extracting a subject or the like. Further, by determining the correction amount from the median value of the luminance histogram, an appropriate correction amount can be obtained.

The invention described in claim 3 is the first or second invention.
In the image processing apparatus described in the above section, by changing the ratio of the highlight portion for changing the exposure correction amount in a scene such as a backlight image or a night person image, a high-definition correction effect according to the scene is obtained. Obtainable.

According to a fourth aspect of the present invention, in the image processing apparatus according to any one of the first to third aspects, further exposure correction is performed on a pattern which is greatly influenced by the dynamic range expansion processing. By examining whether or not it is necessary, excessive exposure correction can be prevented and the atmosphere of the original image can be maintained.

According to a fifth aspect of the present invention, in the image processing apparatus according to any one of the first to fourth aspects, the scene is divided into at least true backlight and others, and if a dynamic range reproduction region is different for each. By doing so, contrast enhancement suitable for the scene can be performed.

According to a sixth aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, when correcting image information biased to a dark level, the shape of the tone curve is low. By using a straight parabolic curve in the region and a convex parabola in the high region, exposure correction can be performed while enhancing the contrast in the low and middle regions. Further, by making the contact point between the straight line and the parabola different according to the scene and the exposure state, it is possible to create an appropriate tone curve.

According to a seventh aspect of the present invention, in the image processing apparatus according to the sixth aspect, since the straight line and the parabola are tangentially connected at the contact point, the tone reproducibility can be maintained at the output level of the contact point.

According to an eighth aspect of the present invention, in the image processing apparatus according to any one of the first to seventh aspects, the exposure correction is performed by changing the exposure correction amount in accordance with the ratio of the exposure suitable portions. This can prevent excessive exposure correction.
Further, the speed can be improved by creating a correction table and performing color conversion processing instead of performing conversion for each pixel.

According to the ninth aspect of the present invention, high-definition correction can be performed by determining the necessity of exposure correction using only a significant highlight portion.

According to the tenth aspect of the present invention, excessive exposure correction can be prevented by changing the amount of exposure correction in accordance with the ratio of the exposure-appropriate portions and performing exposure correction.

According to an eleventh aspect of the present invention, in the program according to the tenth aspect, a highlight portion having a predetermined ratio or more is used by using a level value corresponding to a predetermined ratio of a frequency integrated value in a histogram representing brightness. By judging whether or not there is, it is possible to easily and quickly determine the necessity of exposure correction without performing subject extraction or the like. Further, by determining the correction amount from the median value of the luminance histogram, an appropriate correction amount can be obtained.

According to the twelfth aspect of the present invention, in the program according to the tenth or eleventh aspect, the proportion of the highlight portion for changing the exposure correction amount is determined in a scene such as a backlight image or a night person image. By making them different, a high-definition correction effect according to the scene can be obtained.

The invention according to claim 13 is the invention according to claim 10
12. In the program according to any one of the above 12, it is possible to prevent excessive exposure correction by examining whether or not it is necessary to perform further exposure correction on a pattern having a large degree of influence of the dynamic range expansion processing. The atmosphere of the original picture can be maintained.

The invention according to claim 14 is the invention according to claim 10
In the program according to any one of the thirteenth aspects, it is possible to perform contrast enhancement suitable for the scene by dividing the scene into at least true backlight and others and making the reproduction range of the dynamic range different for each.

The invention described in claim 15 is the invention according to claim 10
In the program according to any one of the fourteenth, when correcting image information that is biased to the dark level, the shape of the tone curve is a straight line in the low region, and a secondary parabola that is upwardly convex in the high region. Exposure correction can be performed while enhancing the contrast in the low and middle regions. Further, by making the contact point between the straight line and the parabola different according to the scene and the exposure state, it is possible to create an appropriate tone curve.

According to a sixteenth aspect of the present invention, in the program according to the fifteenth aspect, the straight line and the parabola are tangentially connected to each other at the contact, so that the gradation reproducibility can be maintained at the output level of the contact.

The invention described in claim 17 is the invention according to claim 10
In the program according to any one of the sixteenth aspects, excessive exposure correction can be prevented by changing the exposure correction amount in accordance with the proportion of the exposure suitable portion and performing the exposure correction. Further, the speed can be improved by creating a correction table and performing color conversion processing instead of performing conversion for each pixel.

According to the eighteenth aspect of the present invention, high-definition correction can be performed by determining the necessity of exposure correction using only a significant highlight portion.

[Brief description of the drawings]

FIG. 1 is a functional block diagram illustrating a schematic configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating an electrical connection of the image processing apparatus.

FIG. 3 is a flowchart illustrating a process performed by the image processing apparatus.

FIG. 4 is an explanatory diagram illustrating a process performed by the image processing apparatus.

FIG. 5 is an explanatory diagram of the same.

FIG. 6 is an explanatory diagram of the same.

FIG. 7 is an explanatory diagram of the same.

FIG. 8 is a flowchart illustrating a process performed by the image processing apparatus.

FIG. 9 is an explanatory diagram for explaining the problem of the present invention.

FIG. 10 is an explanatory diagram of the same.

FIG. 11 is an explanatory diagram illustrating a process performed by the image processing apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 101 Image processing apparatus 102 Dynamic range expansion means 103 Exposure correction means

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C022 AB02 AC69 CA00 5C065 BB00 BB01 BB04 BB12 GG22 GG23 GG31 5C066 AA01 CA07 CA17 EA13 EC01 GA01 HA02 KE01 5C077 LL20 MP01 MP08 NN02 NN03 NP01 PP15 PP32 RR06

Claims (18)

[Claims] 1. An image processing apparatus for correcting image data to an optimum exposure, wherein the image data is corrected in order to enlarge the R, G, B frequency distribution of the image data decomposed in pixel units to a reproduction area. An image processing apparatus comprising: a dynamic range expansion unit; and an exposure correction unit that changes an exposure correction amount according to a ratio of a highlight portion in an image to perform exposure correction. 2. The exposure correction means determines whether or not exposure correction is necessary by using a level value corresponding to a frequency integration value of a predetermined ratio in a histogram representing brightness. The image processing apparatus according to claim 1, wherein the correction amount is determined based on a median of the histogram. 3. The image processing apparatus according to claim 1, wherein the exposure correction unit changes a ratio of a highlight portion for changing the exposure correction amount according to a scene. . 4. The exposure correction means performs exposure correction by changing the exposure correction amount according to the degree of influence of the processing in the dynamic range expansion means and the ratio of a highlight portion. The image processing device according to claim 1. 5. The image processing apparatus according to claim 1, wherein the dynamic range expanding unit changes the correction at least according to a backlight state. 6. The exposure correction means performs tone curve correction as the exposure correction, and the shape of the tone curve in this case is a primary straight line in a low region of the input level and a secondary parabola convex upward in a high region of the input level. 6. The method according to claim 1, wherein the position of the contact is varied according to the scene and the exposure.
The image processing device according to any one of the above. 7. The image processing apparatus according to claim 6, wherein the exposure correction unit connects the two tone curves tangentially at a contact point. 8. The image processing apparatus according to claim 1, wherein the exposure correction unit sets the highlight portion as a continuous area on a two-dimensional position coordinate plane of the image. 9. An image processing apparatus for correcting image data to an optimum exposure, wherein a correction coefficient of the image data is expanded in order to expand a R, G, B frequency distribution of the image data decomposed in pixel units to a reproduction area. A dynamic range expansion correction coefficient calculating means for calculating; an exposure correction coefficient calculating means for calculating a correction coefficient for performing exposure correction by changing an exposure correction amount according to a ratio of a highlight portion in an image; A conversion unit that creates a correction table using the enlargement correction coefficient and the exposure correction coefficient and performs a color conversion process. 10. A computer-readable program for causing a computer to execute image processing for correcting image data to an optimum exposure, wherein a R, G, B frequency distribution of the image data decomposed in pixel units is stored in a reproduction area. A dynamic range expansion process for correcting image data for enlargement, an exposure correction process for changing an exposure correction amount according to a ratio of a highlight portion in an image to perform an exposure correction, and causing a computer to execute. And the program. 11. The exposure correction processing determines whether or not exposure correction is necessary by using a level value corresponding to a frequency integration value of a predetermined ratio in a histogram representing brightness. The program according to claim 10, wherein the correction amount is determined from a histogram median value. 12. The exposure correction process according to claim 10, wherein a ratio of a highlight portion for changing the exposure correction amount is changed according to a scene.
The program according to 1. 13. The exposure correction processing according to claim 1, wherein the exposure correction is performed by changing the exposure correction amount in accordance with a degree of influence of the processing by the dynamic range expansion unit and a ratio of a highlight portion. The program according to claim 10. 14. The dynamic range expansion process according to claim 1,
14. The program according to claim 10, wherein the correction is changed at least according to a backlight state. 15. In the exposure correction processing, tone curve correction is performed as the exposure correction, and the shape of the tone curve in this case is a primary straight line in a low region of the input level and a secondary parabola convex upward in a high region of the input level. 11. The method according to claim 10, wherein the position of the contact is changed according to the scene or the exposure.
15. The program according to any one of items 14 to 14. 16. The program according to claim 15, wherein in the exposure correction processing, the two tone curves are tangentially connected at a contact point. 17. The program according to claim 10, wherein in the exposure correction processing, the highlight part is a continuous area on a two-dimensional position coordinate plane of an image. 18. A computer-readable program for causing a computer to execute image processing for correcting image data to an optimum exposure, wherein a R, G, and B frequency distribution of the image data decomposed in pixel units is stored in a reproduction area. A dynamic range expansion correction coefficient calculation process of calculating a correction coefficient of image data for enlargement, and a correction coefficient for performing exposure correction by changing an exposure correction amount according to a ratio of a highlight portion in an image. A program causing a computer to execute an exposure correction coefficient calculation process and a conversion process of creating a correction table using the dynamic range expansion correction coefficient and the exposure correction coefficient and performing a color conversion process.