JP2009065711A - Image output apparatus, image output method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform correction in lightness so as to clear a region with a local contrast without exceptional application of photometric frame in an input image. <P>SOLUTION: A photometric estimator 350 produces a photometric estimation value of lightness in an imaging signal. A luminance-region producer 361 calculates a dark-side foot value and a light-side foot value in the luminance distribution of an average luminance image in block. A spline generator 362 takes the photometric estimation value as an average luminance value, and further generates a tone curve based on a spline curve using the dark-side and the light-side foot values. A gradation compressor 380 compresses the gradation of the luminance value based on the tone curve. A contrast corrector 390 corrects the contrast of the luminance value compressed in gradation. A gradational correction processor 320 corrects the non-linearly converted gradation of a pixel value in an RGB pixel based on the corrected luminance value in contrast. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image output apparatus, and more particularly to an image output apparatus that performs image brightness correction, a processing method thereof, and a program that causes a computer to execute the method.

  In order to capture or output an image with appropriate brightness, a process for evaluating the brightness of an original image by measurement and calculating a brightness correction amount based on the evaluation value is performed in an imaging device or an image output device. Traditionally done.

  For example, an imaging apparatus is often provided with a mechanism for automatically controlling the operation of an optical system and an image sensor. In particular, a mechanism for controlling the exposure amount to make the brightness of an image captured appropriate is AE. This is called (Auto Exposure). In an imaging apparatus having this AE mechanism, in addition to a system that generates and outputs color image data having an appropriate color and gradation by demosaic, gamma correction, color correction, etc., the brightness of the imaging data is measured and A system for performing exposure control based on the above. That is, a photometric evaluation value obtained by evaluating the brightness of the imaging data based on the measured brightness value of each part of the imaging data is output and compared with a reference value indicating standard brightness. Then, the amount of deviation between the photometric evaluation value and the reference value is set as a difference value, and the difference value approaches “0” in the direction in which the difference value approaches “0”, the sensor charge accumulation time, and the amplifier amplification. The exposure amount is controlled by adjusting the amount and the like.

As such an AE mechanism, it is proposed that an integral value of a luminance value is generated for each small area obtained by dividing captured image data, and a weighted addition of the luminance integral value is used as a brightness evaluation value of the entire image. Has been. For example, there has been proposed an image processing device that divides pixels of the entire screen of image data into a plurality of areas and obtains a photometric value using a central portion of the image as a main area (see, for example, Patent Document 1).
Japanese Patent Laying-Open No. 2004-259177 (FIG. 1)

  In the above-described conventional technology, the photometric value is obtained with the central portion of the image as the main area. However, since subjects vary widely, such as people and landscapes, important subjects are not always located at the center of the screen.

  When the size of the photometric frame is fixed, another object may enter the photometric frame due to the size of the subject. On the other hand, it is actually difficult to change the size and shape of the photometry frame according to the subject.

  In the first place, in order to perform brightness correction using the amount of reflected light as an index, strictly speaking, it is necessary to control in accordance with the reflectance of the subject. That is, a subject with low reflectance should be reproduced darker and a subject with high reflectance should be reproduced brighter. However, in order to know the reflectance of a subject from an input image under an arbitrary illumination, object recognition processing is assumed, which is not realistic from the viewpoint of processing speed.

  Accordingly, an object of the present invention is to evaluate the brightness of an input image by paying attention to a region having a local contrast (contrast). It is another object of the present invention to perform brightness correction processing of the image output device based on the brightness evaluation value of the input image.

  The present invention has been made to solve the above problems, and a first aspect of the present invention is a contrast based on the input luminance which is the luminance of the input image signal for each partial region in the input image signal composed of a plurality of pixels. Is calculated as a local contrast, and a contrast area detecting means for detecting a partial area where the local contrast is greater than a predetermined value as a contrast area, and a bright side brightness of the input brightness and a dark side of the input brightness in the contrast area Equivalent to the boundary luminance value, luminance distribution means for distributing the luminance, boundary luminance value generating means for generating a boundary luminance value of the bright side luminance and the dark side luminance based on the distribution of the bright side luminance and the dark side luminance, and Brightness correction means for performing brightness correction so that the input brightness value is output as a moderate brightness value. The positive means includes a conversion curve calculation means for calculating a conversion curve used for compression of luminance gradation based on the boundary luminance value and the distribution of the input luminance, and luminance of a global luminance image composed of low frequency components of the input image. The overall brightness calculating means for calculating the overall brightness, the tone compression means for compressing the gradation of the input brightness and the brightness of the overall brightness based on the conversion curve, and the slope and the gradation of the conversion curve An image output apparatus comprising contrast correction means for correcting the contrast of a tone-compressed input image composed of the input brightness whose tone is compressed based on the compressed overall brightness, and each procedure realized by each means. An image output method and a program for causing a computer to execute the procedure. This brings about the effect that the input image is corrected based on the boundary luminance value that separates the bright side luminance and the dark side luminance in the contrast region where the luminance contrast is large.

  That is, even if the area of the subject changes, the contrast ratio in the contrast area in the outline does not change, and according to the present invention focusing on the contrast area, the conventional problem that the evaluation value varies depending on the area of the subject is solved. Is done. In addition, according to the present invention, a subject having a relatively low reflectance compared to the surroundings is reproduced darker, and a subject having a relatively high reflectance compared to the surroundings is reproduced brightly. The conventional problem that a subject with a low rate is reproduced unnaturally brightly and a subject with a high reflectance is reproduced unnaturally darkly is solved.

の解である境界輝度値Ｉ_m（但し、Ｅ_G=0＜Ｉ_m＜Ｅ_G=1）を上記境界輝度値として生成してもよい。これは、判別分析手法により境界輝度値を生成させるものである。また、上記２次方程式を簡略化して以下の方程式

によって境界輝度値Ｉ_m（但し、Ｅ_G=0＜Ｉ_m＜Ｅ_G=1）を生成してもよい。この場合、境界輝度値Ｉ_mについての１次方程式として解くことができる。 In this first aspect, the boundary luminance value generation means includes the bright side luminance average value E _{G = 1} , the bright side luminance dispersion value V _{G = 1,} and the dark side luminance average value. E _{G = 0} and the equation based on the dark side luminance dispersion value V _{G = 0}

The boundary luminance value I _m (E _{G = 0} <I _m <E _{G = 1} ) may be generated as the boundary luminance value. This is to generate a boundary luminance value by a discriminant analysis method. In addition, the above quadratic equation is simplified to

The boundary luminance value I _m (where E _{G = 0} <I _m <E _{G = 1} ) may be generated. In this case, it can be solved as a linear equation for the boundary luminance value I _m.

  According to the present invention, the brightness of an input image is evaluated so that a region having a local contrast is clear, and exposure control of the imaging device and brightness correction of the image output device are performed based on the evaluation value. It is possible to achieve an excellent effect that it can be performed.

  Next, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram illustrating a configuration example of a digital still camera 100 according to an embodiment of the present invention. The digital still camera 100 includes a lens 111, an aperture 112, a sensor 113, an amplifier 114, a driver 122, a timing generation unit 123, a camera signal processing unit 130, a photometry unit 140, and a photometry evaluation unit 150. , A camera operation parameter 160, a reference value setting unit 170, a difference value calculation unit 180, and a control unit 190.

  The lens 111 collects a light image (incident light) of the subject. The diaphragm 112 adjusts the amount of light of the optical image collected by the lens 111. Incident light passes through an optical system such as the lens 111 and the diaphragm 112 and reaches the sensor 113.

  The sensor 113 photoelectrically converts a collected light image into an electrical signal, and is realized by, for example, a CCD (Charge Coupled Devices) image sensor. The amplifier 114 amplifies (gains up) the image signal photoelectrically converted by the sensor 113. The image signal amplified by the amplifier 114 is supplied to the camera signal processing unit 130 and the photometry unit 140 via the signal line 119. The driver 122 drives the diaphragm 112 to adjust the aperture amount of the diaphragm. The timing generator 123 adjusts the charge accumulation time by driving the sensor 113.

  The camera signal processing unit 130 performs signal processing such as demosaic processing, gamma correction processing, and color correction processing on the image signal supplied from the amplifier 114 via the signal line 119. The image signal subjected to the signal processing by the camera signal processing unit 130 is output from the signal line 139 and used for recording on a recording medium and display on the display unit.

  The photometry unit 140 outputs the measurement result of the brightness of each part of the image signal supplied from the amplifier 114 via the signal line 119. The measurement result by the photometry unit 140 is supplied to the photometry evaluation unit 150 via the signal line 149.

  The photometric evaluation unit 150 generates a photometric evaluation value that evaluates the brightness of the image signal based on the measurement result supplied from the photometric unit 140 via the signal line 149. The photometric evaluation value generated by the photometric evaluation unit 150 is supplied to the difference value calculation unit 180 via the signal line 159.

  The camera operation parameter 160 is a parameter indicating the current operation state of the digital still camera 100. The reference value setting unit 170 refers to the camera operation parameter 160 and sets a reference value indicating standard brightness. As the reference value, for example, about 18 to 20% converted to 0.4 to 0.5 by a gamma curve can be selected.

  The difference value calculation unit 180 compares the photometry evaluation value generated by the photometry evaluation unit 150 with the reference value set by the reference value setting unit 170, and calculates a difference value as a deviation amount.

  The control unit 190 adjusts the opening amount of the diaphragm 112 of the optical system, the charge accumulation time of the sensor 113, the amplification amount of the amplifier 114, and the like so that the difference value calculated by the difference value calculation unit 180 approaches zero. Is generated. A control signal generated by the control unit 190 is supplied to the diaphragm 112, the sensor 113, the amplifier 114, and the like, and each unit operates so as to obtain an appropriate exposure amount based on the control signal.

  FIG. 2 is a diagram illustrating a configuration example of the photometry unit 140 according to the embodiment of the present invention. The photometry unit 140 includes a luminance calculation unit 141, a non-linear conversion unit 142, a block average generation unit 143, and a block average luminance image memory 144.

  The luminance calculation unit 141 calculates the luminance of each pixel for the input image signal supplied from the amplifier 114 via the signal line 119. Since a normal color sensor can measure only one color for each pixel, in order to measure the brightness of an image, it is necessary to temporarily calculate luminance from such sensor output.

  The non-linear conversion unit 142 performs non-linear conversion on the luminance of each pixel calculated by the luminance calculation unit 141. As this non-linear conversion, for example, a “upwardly convex monotone increasing function” such as logarithmic conversion or gamma correction can be used. This non-linear conversion is performed to preserve the histogram shape for the scaling operation for brightness correction.

  The block average generation unit 143 generates the average of the non-linearly converted luminance as the block average luminance for each block obtained by dividing the image of the image signal nonlinearly converted by the non-linear conversion unit 142 into a lattice shape. That is, it is determined to which block each pixel position belongs, and the integral value of the corresponding block is obtained by adding the luminance subjected to nonlinear transformation, and the integral value is divided by the number of pixels belonging to the block. Thus, the average of the non-linearly transformed luminance in the block is calculated. An image including the blocks divided by the block average generation unit 143 as constituent pixels is referred to as a block average image, and the constituent pixels are referred to as block pixels. That is, the luminance of the block pixel is the block average luminance of the block. One block can adopt a size of about 40 × 30 pixels or 32 × 32 pixels, for example.

  The block average luminance image memory 144 holds the block average image generated by the block average generation unit 143. The block average image held in the block average luminance image memory 144 is supplied as a measurement result by the photometry unit 140 to the photometry evaluation unit 150 via the signal line 149.

  FIG. 3 is a diagram illustrating a generation example of a block average image according to the embodiment of the present invention. FIG. 3A shows each pixel before being divided into blocks. The luminance calculated by the luminance calculation unit 141 is nonlinearly converted by the non-linear conversion unit 142, and is divided into blocks as indicated by the thick lines in FIG.

  Then, the block average generation unit 143 calculates the average of the luminance (nonlinearly converted luminance) belonging to each block as the block average luminance. FIG. 3C shows this state. Thereby, a block average image is generated.

  FIG. 4 is a diagram illustrating a configuration example of the photometric evaluation unit 150 according to the embodiment of the present invention. The photometric evaluation unit 150 includes a contrast area detection unit 151 and a photometric evaluation value generation unit 155. The contrast area detection unit 151 detects, as a contrast area, a partial area having a luminance contrast (brightness difference) greater than a predetermined value for the partial area in the block average image. The photometric evaluation value generator 155 generates a photometric evaluation value of the block average image in the contrast area detected by the contrast area detector 151.

  The contrast region detection unit 151 includes a local contrast calculation unit 152, a threshold determination unit 153, and a contrast region information memory 154.

  The local contrast calculation unit 152 calculates the luminance contrast (luminance difference) for each partial region as the local contrast. In order to calculate the local contrast, for example, the absolute value of the local primary derivative of the luminance or the absolute value of the local secondary derivative in the block average image can be used. Further, an absolute value of a difference between the local maximum value and the local minimum value may be used instead of the differential operator.

  The threshold determination unit 153 compares whether or not the local contrast calculated by the local contrast calculation unit 152 is greater than a predetermined threshold, and determines that the local contrast corresponds to the contrast region if the local contrast is greater than the threshold. Is.

  The contrast area information memory 154 is a memory that holds information as to whether or not a block pixel of the block average image corresponds to a contrast area as contrast area information. For example, corresponding to each block pixel of the block average image, binary data representing “whether or not the block pixel corresponds to the contrast region” is held bit by bit. This contrast area information is supplied to the photometric evaluation value generation unit 155 via the signal line 158.

  The photometric evaluation value generation unit 155 includes a luminance distribution unit 156, a bright side luminance group memory 191, a dark side luminance group memory 192, and a boundary luminance value generation unit 157.

  In the contrast region indicated by the contrast region information supplied via the signal line 158, the luminance distribution unit 156 converts the luminance of each pixel into bright luminance (hereinafter referred to as “bright side luminance”) and dark luminance (hereinafter referred to as “dark side luminance”). )). For example, for each partial area corresponding to the contrast area, the maximum luminance value can be the bright side luminance, and the minimum luminance value can be the dark side luminance. Further, the luminance at the pixel position may be classified into bright side luminance or dark side luminance based on the sign of the local secondary differential of the luminance for each pixel position corresponding to the contrast region. For example, when using a two-dimensional composite of a one-dimensional secondary differential operator [1, -2, 1] as a secondary differential operator applied to an image, the output of the secondary differential operator is If it is negative, the luminance at the pixel position is the bright side luminance, and if it is positive, the luminance is the dark side luminance. Incidentally, in the field of image processing, the above-described example of the second-order differential operator, in which the sign is inverted, is also commonly used as the second-order differential operator. Must be reversed from the above.

  Note that when the local contrast is obtained from the absolute value of the local first derivative of the luminance in the contrast region detection unit 151 or when the local contrast is obtained from the absolute value of the difference between the local maximum value and the minimum value, the luminance distribution is performed. It is suitable for the unit 156 to distribute the luminance from the maximum value and the minimum value of the luminance. On the other hand, when the local contrast is obtained from the absolute value of the local second derivative of the luminance in the contrast region detecting unit 151, the luminance distribution unit 156 uses the intermediate result to obtain the luminance by the sign of the local second derivative of the luminance. Is efficient to distribute. However, these combinations are not particularly limited.

  The bright-side luminance group memory 191 holds the luminance (non-linearly converted luminance) distributed to the bright-side luminance by the luminance distribution unit 156. The dark side luminance group memory 192 holds the luminance (the luminance subjected to nonlinear conversion) distributed to the dark side luminance by the luminance distribution unit 156.

  The boundary luminance value generation unit 157 determines the bright side luminance and the dark side luminance based on the respective distributions of the bright side luminance held in the bright side luminance group memory 191 and the dark side luminance held in the dark side luminance group memory 192. The boundary luminance value is generated as a photometric evaluation value. That is, a boundary luminance value that can separate the bright side luminance and the dark side luminance as well as possible is generated. The boundary luminance value generated by the boundary luminance value generation unit 157 is output via the signal line 159 as a photometric evaluation value.

In the embodiment of the present invention, a discriminant analysis method is used as follows to generate the boundary luminance value. That is, when a certain luminance value I is given, the luminance value I is assigned to a group having a larger posterior probability (posterior) as to whether it belongs to the dark side luminance group or the bright side luminance group. It is assumed that the random variable of which group belongs to G, G = 0 for the dark side luminance group, and G = 1 for the bright side luminance group. Here, the ratio Λ of posterior probabilities belonging to each group is given by the following equation.

  Taking the logarithm of this ratio Λ, it can be determined that if log Λ <0, the luminance value I belongs to the dark side luminance group, and if log Λ> 0, the luminance value I belongs to the bright side luminance group. If logΛ = 0, the luminance value I is on the boundary luminance that separates both groups. In this case, the base of the logarithm may be any, and a common logarithm or a natural logarithm may be selected as appropriate.

  p (I | G) indicates the luminance distribution of the dark side luminance group (when G = 0) and the bright side luminance group (when G = 1). Here, in order to apply discriminant analysis, it is assumed that the distribution on the luminance axis of the dark side luminance group and the bright side luminance group obtained from the contrast area in the block average image follows a normal distribution.

p (G) indicates the probability of belonging to which group regardless of the luminance value. In the embodiment of the present invention, attention is paid to the dark side brightness and the bright side brightness of the contrast region, so the number of data belonging to both groups is substantially the same, and p (G = 0) = p (G = It can be considered as 1). Therefore, the following equation is obtained for the boundary luminance value I _m to separate the two groups.

However, the average value of bright side luminance is E _{G = 1} , the variance value of bright side luminance is V _{G = 1} , the average value of dark side luminance is E _{G = 0} , and the variance value of dark side luminance is V _{G = 0.} Yes. This equation are the quadratic equation for the boundary luminance value I _m, the solution there are two. Of these two solutions, the boundary luminance value _Im is obtained by selecting the _{one in} the range of E _{G = 0} <I _m <E _{G = 1} .

Here, as an example of discriminant analysis, secondary discriminant analysis has been described as an example, but other than this, for example, linear discriminant or support vector machine may be used. For example, in the case of linear discrimination, the above equation is simplified as follows. This equation is a linear equation for the boundary luminance value I _m (where E _{G = 0} <I _m <E _{G = 1} ).

  FIG. 5 is a diagram illustrating a generation example of the boundary luminance value in the embodiment of the present invention. Here, five sets of bright-side luminance and dark-side luminance in the contrast region are displayed. That is, the first set of bright side luminance 511 and dark side luminance 521, the second set of bright side luminance 512 and dark side luminance 522, the third set of bright side luminance 513 and dark side luminance 523 , A fourth set of bright side luminance 514 and dark side luminance 524, and a fifth set of bright side luminance 515 and dark side luminance 525. These sets are groups that indicate the maximum and minimum values of luminance in the luminance distribution unit 156, or are distributed according to the sign of the local second derivative of luminance.

  A distribution graph above the distribution of these sets is a bright side luminance group 510 and a dark side luminance group 520. This distribution graph shows the luminance obtained by nonlinear transformation in the horizontal direction (in the example of this figure, the logarithm of luminance L), and the luminance is brighter toward the right side. In addition, the existence probability of each luminance is shown in the upper direction, and the existence probability is higher toward the upper side.

  As described above, the boundary luminance value is obtained by finding the solution of the equation (Equation 2). This boundary luminance value separates the bright side luminance group 510 and the dark side luminance group 520.

  FIG. 6 is a diagram showing a generation example of the bright side luminance group and the dark side luminance group in the embodiment of the present invention. FIG. 6A is an example of a histogram before the bright side luminance and the dark side luminance of the luminance subjected to nonlinear conversion are distinguished. By obtaining the boundary luminance value for an image showing such a distribution, the distributions as shown in FIGS. That is, FIG. 5B is a histogram example of a bright side luminance group of non-linearly converted luminance, and FIG. 10C is a histogram example of a dark side luminance group of non-linearly converted luminance.

  In these histogram examples, normalization is performed by the maximum value of each existence probability.

  FIG. 7 is a diagram illustrating an example of a control procedure of the digital still camera 100 according to the embodiment of the present invention. First, when incident light from a subject passes through an optical system such as the lens 111 and the diaphragm 112 and is imaged by the sensor 113 (step S901), the photometric unit 140 performs photometric processing on the input image amplified by the amplifier 114. Performed (step S910). Then, based on the measurement result of the photometric process, the photometric evaluation unit 150 performs the photometric evaluation process (step S920).

  When the photometric evaluation value obtained by the photometric evaluation process is output (step S902), the difference value calculation unit 180 generates a difference value from the reference value set by the reference value setting unit 170 (step S903). Based on the generated difference value, the control unit 190 generates control signals in the driver 122, the timing generation unit 123, the amplifier 114, and the like to perform operation control (step S904).

  FIG. 8 is a diagram showing an example of a processing procedure of the photometric process (step S910) in the embodiment of the present invention. When an input image is input via the signal line 119 (step S911), the following processing is performed for each pixel of the input image (loop L991). Here, each pixel of the input image is represented by two-dimensional coordinates (p, q). However, p and q are both integers of 0 or more.

  First, the luminance of the input pixel (p, q) is calculated by the luminance calculation unit 141 (step S912). The calculated luminance is nonlinearly converted by the nonlinear converter 142 (step S913).

  Then, in the block average generation unit 143, the block to which the input pixel (p, q) belongs is determined (step S914), and the variable of the integrated value of the determined block is stored in the variable of the input pixel (p, q). The non-linearly converted luminance is added (step S915).

  When the above-described processing is performed for all input pixels (p, q) in the loop L991, the following processing is performed for each block (loop L992). That is, a process of dividing the integral value of each block by the number of pixels of that block is performed (step S916). Thereby, the division result is held in the block average luminance image memory 144 as the block average luminance of each block (step S917).

  FIG. 9 is a diagram showing an example of a processing procedure of the photometric evaluation process (step S920) in the embodiment of the present invention. When a block average image is input via the signal line 149 (step S921), the contrast area detection unit 151 performs a contrast area detection process on the block average image (step S930). Further, for the contrast area detected by the contrast area detection unit 151, a photometric evaluation value generation process is performed in the photometric evaluation value generation unit 155 (step S940). The generated photometric evaluation value is output via the signal line 159 (step S925).

  FIG. 10 is a diagram showing an example of the processing procedure of the contrast area detection processing (step S930) in the embodiment of the present invention. When the block average image is input via the signal line 149 (step S931), the following processing is performed for each pixel of the block average image (loop L993). Here, each pixel of the block average image is represented by two-dimensional coordinates (r, s). However, r and s are both integers of 0 or more.

  First, the local contrast calculation unit 152 calculates the luminance local contrast at the pixel (r, s) of the block average image (step S932). As described above, in order to calculate the local contrast, for example, the absolute value of the local primary derivative or the absolute value of the local secondary derivative of the luminance in the block average image can be used.

  If the generated local contrast is larger than the threshold value (step S933), contrast area information indicating that the pixel (r, s) corresponds to the contrast area is stored in the contrast area information memory 154 by the threshold determination unit 153. (Step S934). The contrast area information is supplied to the luminance distribution unit 156 of the photometric evaluation value generation unit 155 via the signal line 158.

  FIG. 11 is a diagram illustrating an example of a processing procedure of the photometric evaluation value generation process (step S940) according to the embodiment of the present invention. When the block average image is input via the signal line 149 (step S941) and the contrast area information is input via the signal line 158 (step S942), the luminance distribution unit 156 changes the brightness to the bright side luminance or the dark side luminance. A luminance distribution process is performed (step S950). Then, the boundary luminance value generation processing is performed by the boundary luminance value generation unit 157 based on the distribution of the bright side luminance and the dark side luminance (step S960). The boundary luminance value generated by the boundary luminance value generation processing is output as a photometric evaluation value via the signal line 159 (step S945).

  FIG. 12 is a diagram showing an example of the processing procedure of the luminance distribution processing (step S950) in the embodiment of the present invention. When a block average image is input via the signal line 149 (step S951) and contrast area information is input via the signal line 158 (step S952), the following processing is performed on each pixel of the block average image (step S952). Loop L995). Here, each pixel of the block average image is represented by two-dimensional coordinates (r, s). However, r and s are both integers of 0 or more.

  First, if the contrast region information indicates that the pixel (r, s) of the block average image is in the contrast region (step S953), the luminance distribution unit 156 changes the luminance of the pixel (r, s) to the bright side luminance. Or it is distributed to the dark side luminance. In the example of the figure, the sign of the local second derivative of luminance is used as a distribution reference. That is, the local secondary derivative of the luminance at the pixel (r, s) is calculated (step S954), and if the calculated local secondary derivative is smaller than “0” (step S955), the value of the local secondary derivative is calculated. Is stored in the bright side luminance group memory 191 as the bright side luminance (step S956). On the other hand, if the calculated local secondary derivative is not smaller than “0” (step S955), the value of the local secondary derivative is held in the dark side luminance group memory 192 as the dark side luminance (step S957).

FIG. 13 is a diagram showing an example of the processing procedure of the boundary luminance value generation processing (step S960) in the embodiment of the present invention. When the bright side luminance and the dark side luminance are input from the bright side luminance group memory 191 and the dark side luminance group memory 192 (step S961), an average value EG _{= 1 of the} bright side luminance is calculated (step S962). The side luminance variance value V _{G = 1} is calculated (step S963). Similarly, an average value EG _{= 0} of the dark side luminance is calculated (step S964), and a variance value V _{G = 0 of} the dark side luminance is calculated (step S965).

Then, the solution of the quadratic equation are calculated for the following boundary luminance value I _m (step S966).

Here, since there are two solutions because of the quadratic equation, the solution that is in the range of E _{G = 0} <I _m <E _{G = 1} is selected as the solution (step S967). The selected solution is output as a boundary luminance value (that is, a photometric evaluation value) via the signal line 159 (step S968).

  As described above, according to the embodiment of the present invention, the luminance of each pixel is distributed to the bright side luminance or the dark side luminance by the photometric evaluation value generation unit 155 in the contrast region detected by the contrast region detection unit 151. A boundary luminance value is generated as a photometric evaluation value based on the distribution of the bright side luminance and the dark side luminance. This photometric evaluation value is compared with a reference value in the difference value calculation unit 180, and can be used for brightness correction operation control in the control unit 190.

  In the above-described embodiment, an example is shown in which the present invention is applied to brightness correction in a digital still camera, but the present invention is not limited to this. For example, the present invention can be applied to brightness correction of an output image in a computer system as follows.

  FIG. 14 is a diagram showing a configuration example of the computer system 200 according to the embodiment of the present invention. The computer system 200 includes a computer including a processor 211, a memory 212, a display controller 213, and an input / output interface 214.

  The processor 211 is an arithmetic device that executes a program and controls each device. The memory 212 is a place for storing temporary data necessary for executing such a program and controlling each device.

  The display controller 213 is an interface for connecting a display and incorporates a graphics controller. A display 220 is connected to the display controller 213. The display 220 is a display device that displays an output image from the computer 210, and is realized by, for example, an LCD (Liquid Crystal Display).

  The input / output interface 214 is an interface for connecting peripheral devices, and incorporates a USB (Universal Serial Bus) controller and the like. For example, a mouse 230, a keyboard 240, a recording medium access device 250, a printer 260, and the like are connected to the input / output interface 214. The mouse 230 receives a user instruction by an operation such as click or drag. The keyboard 240 accepts user instructions such as character keys and numeric keys. The recording medium access device 250 is a device that performs writing and reading with respect to an inserted recording medium. The printer 260 is a printing device that prints an output image from the computer 210.

  In such a computer system 200, a printing system is assumed in which image data recorded on a recording medium inserted into the recording medium access device 250 is automatically read and an image that has been appropriately corrected is printed by the printer 260. . This printing system may be realized by a program operating on the processor 211 or may be controlled by hardware.

  FIG. 15 is a diagram showing a display example by the printing system in the embodiment of the present invention. In this display example, the outermost frame 500 represents a GUI (Graphical User Interface) display range by the printing system, and this is usually displayed on the entire screen of the display 220.

  The area occupying most of the frame 500 is divided into 4 × 4 clickable small windows 510. The small window 510 is an area for displaying a thumbnail of an image read from the recording medium by the recording medium access device 250. Since the number of images recorded on the recording medium is not necessarily 16 or less, the thumbnail display area can be scroll-displayed by the scroll bar 531 and the two scroll buttons 532 and 533.

  Thick frames 521 and 522 displayed around the thumbnail are indicators that indicate that the image is in a selected state. When each thumbnail is clicked, the selected or unselected state of the image is switched (toggled). When the print button 550 in the upper right corner of the screen is clicked, automatic correction and printing of the currently selected image is started.

  FIG. 16 is a diagram illustrating an example of a processing procedure of the printing system according to the embodiment of the present invention. When the recording medium is inserted into the recording medium access device 250, image data for display is read from the recording medium (step S801), and thumbnail images are displayed side by side on the GUI display screen (step S802). Thereafter, while the recording medium is inserted into the recording medium access device 250, the following processing is repeated (loop L896).

  When an operation input by the user is accepted (step S803), the type of the operation input is determined. If the operation input is a thumbnail click operation (step S804), the selected or unselected state of the image is switched according to the current state of the corresponding image (step S805). That is, if the current state is the selected state, the selected state is canceled (step S806), or if the current state is not the selected state, the selected state is set (step S807), and the operation for the current operation input is completed. To do.

  If the operation input is an operation of the scroll bar 531 or the scroll buttons 532 and 533, the thumbnail display area is scrolled to complete the operation for the current operation input.

  If the operation input is a click operation on the print button 550, brightness correction processing is performed on all the images that are currently selected as target images (step S872), and the brightness correction is performed. The image subjected to is printed by the printer 260 (step S873). When printing is completed for all target images in the selected state (loop L897), the operation for the current operation input is completed.

  When the operation for one operation input is completed, it is determined whether or not a recording medium is inserted into the recording medium access device 250, and if it is inserted, the above processing is repeated. On the other hand, when it is determined that the recording medium is not inserted at the time when the operation for one operation input is completed, the series of processes is ended.

  FIG. 17 is a diagram illustrating a functional configuration example of brightness correction processing of the computer system 200 according to the embodiment of the present invention. In this brightness correction process, brightness correction is performed on the red component R, green component G, and blue component B of the pixels of the RGB image input through the signal lines 307 to 309, and the result is output to the signal lines 337 to 339. . This brightness correction processing includes a conversion unit 310, a gradation correction processing unit 320, an inverse conversion unit 330, a photometry unit 340, a photometry evaluation unit 350, a luminance range generation unit 361, a spline generation unit 362, The tone curve memory 363, the γ_comp parameter memory 364, the luminance generation unit 371, the nonlinear conversion unit 372, the interpolation unit 377, the gradation compression unit 380, and the contrast correction unit 390 are provided.

  Note that the configuration of the brightness correction processing of FIG. 17 described in the embodiment of the present invention is not an essential configuration requirement in the present invention. That is, the brightness for which the brightness correction is performed so that the input luminance value corresponding to the photometric evaluation value obtained based on the photometric evaluation value calculation method in the embodiment of the present invention described above is converted into a moderate output luminance. If there is a function of brightness correction processing, image display with brightness correction using the present invention is possible. The brightness correction processing corresponding to the configuration other than the configuration example of FIG. 17 includes a method for controlling the luminance scaling coefficient, a method for controlling the luminance bias value, a method for controlling the power exponent of the gamma curve with the photometric evaluation value, or Combinations of these are possible. Description of the detailed contents of these processes is omitted.

  The conversion unit 310 performs non-linear conversion on pixel values of each component of the RGB image input through the signal lines 307 to 309. The conversion unit 310 includes non-linear conversion units 311 to 313, and performs non-linear conversion on the pixel values of the red component R, the green component G, and the blue component B, respectively, and outputs them to the signal lines 317 to 319. Here, for example, logarithmic transformation or gamma correction can be applied as the nonlinear transformation.

  The gradation correction processing unit 320 corrects (converts) the gradation of pixel values (pixel values of RGB images subjected to nonlinear conversion) input through the signal lines 317 to 319. The gradation correction processing unit 320 includes gradation correction units 321 to 323, and corrects the gradations of the pixel values of the red component R, the green component G, and the blue component B, respectively, to the signal lines 327 to 329. Output. Note that the non-linearly converted luminance value (signal line 375) and the contrast-corrected luminance value (signal line 399) used in the gradation correction processing unit 320 will be described later.

  The inverse conversion unit 330 performs nonlinear inverse conversion on the pixel values (pixel values of the RGB image subjected to gradation correction) input through the signal lines 317 to 319. The inverse transform unit 330 includes nonlinear inverse transform units 331 to 333, which nonlinearly transform the pixel values of the red component R, the green component G, and the blue component B, respectively, and output them to the signal lines 337 to 339.

  The photometry unit 340 is the same as the photometry unit 140 described with reference to FIG. 2, and brightness measurement results (that is, block average luminance images) for the pixel values of each component of the RGB image input through the signal lines 307 to 309. ) To the signal line 349.

  The photometric evaluation unit 350 is the same as the photometric evaluation unit 150 described with reference to FIG. 4, and evaluated the brightness of the image signal based on the measurement result supplied from the photometric unit 340 via the signal line 349. A photometric evaluation value is generated and output to the signal line 359.

In the luminance distribution of the block average luminance image supplied from the photometry unit 340 via the signal line 349, the luminance region generation unit 361 has a ratio of the number of pixels with luminance equal to or less than a predetermined boundary value in the total number of pixels. The ratio of the number of pixels with a luminance equal to or higher than a predetermined boundary value to the total number of pixels and the dark side base value L _dark ^(nl) which is a boundary value that becomes a value of 0.5 (for example, 0.5) is almost a predetermined value The _{bright side} base value L _bright ^(nl) , which is a boundary value that becomes (for example, 0.5), is calculated. Note that “(nl)” added to the upper right of each value indicates that the value has been nonlinearly transformed.

  Note that the photometric evaluation unit 350 and the luminance range generation unit 361 read out the noise level and saturation level of the luminance value from an internal memory (not shown), and remove the luminance value below the noise level and the luminance value above the saturation level. You may make it do.

The spline generation unit 362 treats the photometric evaluation value supplied from the photometric evaluation unit 350 through the signal line 359 as the average luminance value L _average ^(nl) , and further, the dark side skirt value L _dark ⁽ supplied from the luminance region generation unit 361 ^{). The} spline curve is generated using ^nl) and the bright side base value L _bright ^(nl) . This spline curve represents a tone curve (conversion curve) for tone compression. The tone curve is calculated based on the luminance distribution of the image so that the input image is reproduced with appropriate brightness. A look-up table representing the tone curve is held in the tone curve memory 363. In addition, a γ_comp parameter representing the shape (inclination) of the tone curve is held in the γ_comp parameter memory 364.

The luminance generation unit 371 calculates a luminance value L corresponding to the pixel position from the pixel value of each component of the RGB image input through the signal lines 307 to 309. The non-linear conversion unit 372 performs non-linear conversion on the luminance value generated by the luminance generation unit 371. An image composed of the luminance value L ^(nl) nonlinearly transformed by the nonlinear transformation unit 372 is supplied to the interpolation unit 377, the gradation compression unit 380, and the gradation correction processing unit 320 via the signal line 375.

The interpolating unit 377 interpolates the block average luminance image supplied from the photometric unit 340 and enlarges it to have the same pixel value as the image supplied from the non-linear converting unit 372 via the signal line 375. The luminance value of the pixels constituting the enlarged image is supplied to the gradation compression unit 380 as a general luminance value L _l ^(nl) via the signal line 379. That is, this overall luminance value has the low frequency component of the image supplied from the nonlinear conversion unit 372.

The gradation compression unit 380 compresses the gradation of the luminance value L ^(nl) and the overall luminance value L _l ^(nl) . The tone compression unit 380 includes two mapping units 381 and 382. The mapping unit 381 compresses the gradation of the luminance value L ^(nl) supplied via the signal line 375 based on the tone curve read out from the tone curve memory 363, and compresses the compressed luminance value L _c ^{(nl )} . The mapping unit 382 compresses the gradation of the overall luminance value L _l ^(nl) supplied via the signal line 379 based on the tone curve read from the tone curve memory 363, and compresses the compressed overall luminance value L _cl ^(nl) is generated. Such tone compression based on the tone curve is called tone curve correction.

Based on the γ_comp parameter stored in the γ_comp parameter memory 364 and the compressed overall luminance value L _cl ^(nl) supplied from the mapping unit 382, the contrast correction unit 390 compresses the compressed luminance value L supplied from the mapping unit 381. The contrast of the image composed of _c ^(nl) is corrected and output as a luminance value L _u ^(nl) via the signal line 399.

Specifically, in the contrast correction unit 390, based on the compressed luminance value L _c ^(nl) and the compressed overall luminance value L _cl ^(nl) in which the gradation is compressed, and the gain value g, A contrast-corrected luminance value L _u ^(nl) is calculated.
L _u ^(nl) = g · (L _c ^(nl) −L _cl ^(nl) ) + L _c ^(nl)
However, the gain value g is a value depending on the γ_comp parameter.

Note that the image formed of the luminance value ^{_{(L c (nl) -L cl}} (nl)) from the image consisting of brightness value L _c ^(nl), the very low frequency region of an image made up of the brightness value L _c ^(nl) This is a subtraction of the overall luminance image made up of components. Therefore, the image composed of the luminance value L _u ^(nl) is an image in which the frequency component excluding the extremely low frequency range of the image composed of the luminance value L _c ^(nl) is emphasized by the gain value g.

The luminance value L _u ^(nl) subjected to contrast correction by the contrast correction unit 390 is supplied to the gradation correction units 321 to 323 via the signal line 399. Accordingly, the gradation correction processing unit 320 corrects the gradation of the pixel values R ^(nl) , G ^(nl) , or B ^(nl) supplied from the signal lines 317 to 319 by the following formula. The pixel value R _u ^(nl) , G _u ^(nl) or B _u ^(nl) is calculated.
R _u ^(nl) = chromagain · (R ^(nl) −L ^(nl) ) + L _u ^(nl)
G _u ^(nl) = chromagain · (G ^(nl) −L ^(nl) ) + L _u ^(nl)
B _u ^(nl) = chromagain · (B ^(nl) −L ^(nl) ) + L _u ^(nl)
However, chromagain is a coefficient of a predetermined value for adjusting the saturation of each component of R, G, and B.

That is, the pixel value R _u ^(nl) , G _u ^(nl) , or B _u ^(nl) is an image and gradation composed of the pixel value R ^(nl) , G ^(nl) , or B ^(nl) , respectively. A pixel of an image obtained by multiplying each pixel value of a difference image, which is a difference from an image composed of a luminance value L ^(nl) before conversion, by chromagain and adding an image composed of a luminance value L _u ^(nl) after contrast correction Value. Thereby, the luminance value after gradation conversion is reflected in the RGB value subjected to nonlinear conversion.

The pixel values R _u ^(nl) , G _u ^(nl) , and B _u ^(nl) generated in this way are nonlinearly inversely transformed by the nonlinear inverse transformers 331 to 333, respectively, and are transmitted to the signal lines 337 to 339, respectively. Pixel values R _u , G _u , or B _u are output.

  That is, the brightness correction process according to the example of FIG. 17 is a combination of the tone curve correction process based on the tone curve calculated by examining the brightness of the input image and the contrast correction process for recovering the deterioration of contrast caused thereby. Become.

  FIG. 18 is a diagram showing an example of a tone curve generated by the spline generator 362 according to the embodiment of the present invention. The horizontal axis direction of this graph represents a value (for example, logarithmic value) of the input luminance before gradation correction, and the vertical axis direction is a nonlinear conversion of the output luminance after gradation correction by the tone curve CL. It represents a value (for example, a logarithmic value).

  In the spline generation unit 362, nine control points P1 to P9 are set. The spline generator 362 calculates coordinates on a cubic spline curve that interpolates between the control points P1 to P9, and generates a look-up table for the tone curve CL.

The control point P1 is set to a point where the input luminance is a predetermined minimum level and the output luminance is a predetermined minimum level L _base ^(nl) . The control point P2 has a predetermined noise level L _noise ^(nl) that is a luminance at which the input luminance can be regarded as a noise level, and is set to a point at which the output luminance becomes the minimum level L _base ^(nl) . The control point P3 is set to a point where the input luminance becomes a luminance value twice the noise level L _noise ^(nl) and the output luminance becomes the minimum level L _base ^(nl) .

The control point P4 is set to a point at which the input luminance becomes the dark side base value L _dark ^(nl) and the output luminance becomes the luminance value _Lankle ^(nl) which is a luminance value of a substantially black level. The control point P5 is set to a point where the input luminance is a luminance value twice as large as the dark side base value L _dark ^(nl) and the output luminance is a luminance value twice as large as the luminance value _Lankle ^(nl) . The control point P6 is a point at which the input luminance is an average level (average luminance value) L _average ^{(nl) of} the input luminance and the output luminance is a predetermined almost intermediate intermediate luminance level L _mid ^(nl) in the luminance range of the output luminance. Set to In the control point P7, the input luminance is a half luminance value of the _bright side base value L _bright ^(nl) , and the output luminance is a half luminance value L _shoulder ^(nl) which is a luminance value of almost white level. It is set to the point that becomes the luminance value. The control point P8 is set to a point where the input luminance becomes the bright side base value L _bright ^(nl) and the output luminance becomes the luminance value L _shoulder ^(nl) . The control point P9 is set to a point where the input luminance becomes the maximum value of the predetermined input luminance and the output luminance becomes the maximum value of the predetermined output luminance.

  That is, the control points P1, P2, P3, and P9 are given fixed values in advance according to the noise level, saturation level, gradation characteristics, and the like of the output device assumed to be the input image, and change according to the luminance distribution of the input image. There is no. The coordinates of the control points P4, P6, and P8 in the vertical axis direction are similarly fixed, but the coordinates in the horizontal axis direction are calculated from the luminance distribution of the input image. Further, the positions of the control points P5 and P7 are subordinately determined when the positions of the control points P4, P6, and P8 are determined by the input image. The slope of the line segment connecting the control points P5 and P7 is the γ_comp parameter held in the γ_comp parameter memory 364.

  Since the control point P6 is mapped to the moderate brightness of the output image, it is the most important control point in determining the brightness of the reproduced image. Therefore, in the embodiment of the present invention, the photometric evaluation value generated by the photometric evaluation unit 350 is used as the coordinate (average luminance value) in the horizontal axis direction of the control point P6, so that an image can be obtained with more appropriate brightness. To be reproduced.

  19 and 20 are diagrams illustrating an example of the procedure of the brightness correction process of the printing system according to the embodiment of the present invention. In this brightness correction processing, the entire input image is processed in a two-pass configuration that scans twice. The first pass performs block average image generation and tone curve calculation using photometric evaluation, and is shown in FIG. The second pass performs gradation correction processing for all pixels based on the block average image and the tone curve, and is shown in FIG.

  In this brightness correction process, first, the brightness of the target image is measured by the photometry unit 340 as a photometry process, and a block average luminance image is generated (step S810). The procedure in this photometry process is the same as the photometry process (step S910) described in FIG.

  Based on the generated block average luminance image, a photometric evaluation value is generated by the photometric evaluation unit 350 as a photometric evaluation process (step S820). The procedure in this photometric evaluation process is the same as the photometric evaluation process (step S920) described in FIG.

  In addition, the luminance range generation unit 361 generates the base values on the bright side and the dark side of the luminance histogram from the luminance distribution of the block average luminance image (step S874).

Then, in the spline generation unit 362, the photometric evaluation value supplied from the photometric evaluation unit 350 is set as the average luminance value L _average ^(nl), and the bright side and dark side base values supplied from the luminance range generation unit 361 are set to be bright. A tone curve is generated as the side base value L _bright ^(nl) and the dark side base value L _dark ^(nl) (step S875). A look-up table representing the generated tone curve is held in the tone curve memory 363. Further, the slope of the line segment connecting the tone curve control points P5 and P7 is generated as the γ_comp parameter (step S876). The γ_comp parameter is held in the γ_comp parameter memory 364.

  Next, the following processing is performed for each pixel of the input target image (loop L898). Here, each pixel of the target image is represented by two-dimensional coordinates (p, q). However, p and q are both integers of 0 or more.

First, when the RGB value of the pixel (p, q) corresponding to the target image is input (step S880), the luminance generation unit 371 generates the luminance value L of the pixel (p, q) (step S881). The luminance value L is nonlinearly converted by the nonlinear conversion unit 372 and output as the luminance value L ^(nl) (step S882).

Then, the interpolation processing of the block average luminance image generated in the first pass is performed by the interpolation unit 377 based on the non-linearly converted luminance value L ^(nl) , and the overall luminance corresponding to the pixel (p, q). A value L _l ^(nl) is generated (step S883).

The luminance value L ^(nl) and the overall luminance value L _l ^(nl) of the pixel (p, q) generated in this way are corrected by the tone curve correction (gradation) based on the tone curve by the mapping unit 381 or 382, respectively. Compression) is performed (step S884). As a result, a compressed luminance value L _c ^(nl) and a compressed global luminance value L _cl ^(nl) are generated. The contrast correction unit 390 corrects the contrast of the compressed luminance value L _c ^(nl) based on the compressed overall luminance value L _cl ^(nl) and the γ_comp parameter (step S885). The luminance value L _u ^(nl) after the contrast correction is used for the gradation correction process.

On the other hand, the RGB values of the corresponding pixel (p, q) of the target image are nonlinearly transformed by the nonlinear transformation units 311 to 313 (step S886). The non-linearly transformed RGB values (R ^(nl) , G ^(nl) and B ^(nl) ) are based on the non-linearly transformed luminance value L ^(nl) and the luminance value L _u ^(nl) after contrast correction. Gradation correction is performed by the gradation correction units 321 to 323 (step S886). The RGB values (R _u ^(nl) , G _u ^(nl), and B _u ^(nl) ) subjected to gradation correction are nonlinearly inversely transformed by the nonlinear inverse transform units 331 to 333 (step S888). Thereby, the RGB values (R _u , G _u and B _u ) subjected to nonlinear inverse transformation are output (step S889).

  When these processes are performed for all pixels, the brightness correction process for the target image ends.

  Thus, the present invention can be widely applied not only to brightness correction in an imaging apparatus such as a digital still camera but also to brightness correction during printing in a computer system.

  The embodiment of the present invention is an example for embodying the present invention and has a corresponding relationship with the invention-specific matters in the claims as shown below, but is not limited thereto. However, various modifications can be made without departing from the scope of the present invention.

  That is, in claim 1, the contrast area detecting means corresponds to the contrast area detecting unit 151, for example. The luminance distribution unit corresponds to the luminance distribution unit 156, for example. The boundary luminance value generation unit corresponds to the boundary luminance value generation unit 157, for example. Further, the brightness correction unit may include, for example, a spline generation unit 362, an interpolation unit 377, a contrast correction unit 390, and the like. The conversion curve calculation means corresponds to the spline generation unit 362, for example. The overall luminance calculation means corresponds to the interpolation unit 377, for example. A gradation compression unit corresponds to the gradation compression unit 380, for example. A contrast correction unit corresponds to the contrast correction unit 390, for example.

  Further, in claims 4 and 5, the contrast region detection procedure corresponds to, for example, step S930. The luminance distribution procedure corresponds to, for example, step S950. The boundary luminance value generation procedure corresponds to, for example, step S960. The brightness correction procedure can include, for example, steps S875 and S883 to S885. The conversion curve calculation procedure corresponds to, for example, step S875. The overall luminance calculation procedure corresponds to step S883, for example. The gradation compression procedure corresponds to step S884, for example. The contrast correction procedure corresponds to step S885, for example.

  The processing procedure described in the embodiment of the present invention may be regarded as a method having a series of these procedures, and a program for causing a computer to execute these series of procedures or a recording medium storing the program May be taken as

It is a figure which shows one structural example of the digital still camera 100 in embodiment of this invention. It is a figure which shows the example of 1 structure of the photometry part 140 in embodiment of this invention. It is a figure which shows the example of a production | generation of the block average image in embodiment of this invention. It is a figure which shows the example of 1 structure of the photometry evaluation part 150 in embodiment of this invention. It is a figure which shows the example of a production | generation of the boundary luminance value in embodiment of this invention. It is a figure which shows the example of a production | generation of the bright side luminance group and dark side luminance group in embodiment of this invention. It is a figure which shows an example of the control procedure of the digital still camera 100 in embodiment of this invention. It is a figure which shows an example of the process sequence of the photometry process in embodiment of this invention. It is a figure which shows an example of the process sequence of the photometry evaluation process in embodiment of this invention. It is a figure which shows an example of the process sequence of the contrast area | region detection process in embodiment of this invention. It is a figure which shows an example of the process sequence of the photometric evaluation value production | generation process in embodiment of this invention. It is a figure which shows an example of the process sequence of the brightness | luminance distribution process in embodiment of this invention. It is a figure which shows an example of the process sequence of the boundary luminance value production | generation process in embodiment of this invention. It is a figure which shows the example of 1 structure of the computer system 200 in embodiment of this invention. It is a figure which shows the example of a display by the printing system in embodiment of this invention. It is a figure which shows an example of the process sequence of the printing system in embodiment of this invention. It is a figure which shows the function structural example of the brightness correction process of the computer system 200 in embodiment of this invention. It is a figure which shows an example of the tone curve produced | generated by the spline generation part 362 by embodiment of this invention. It is a figure which shows the example of a procedure of the first half of the brightness correction process of the printing system in embodiment of this invention. It is a figure which shows the example of a procedure of the latter half of the brightness correction process of the printing system in embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Digital still camera 111 Lens 112 Aperture 113 Sensor 114 Amplifier 122 Driver 123 Timing generation part 130 Camera signal processing part 140,340 Photometry part 141 Luminance calculation part 142 Nonlinear conversion part 143 Block average production | generation part 144 Block average luminance image memory 150,350 Photometric evaluation unit 151 Contrast region detection unit 152 Local contrast calculation unit 153 Threshold determination unit 154 Contrast region information memory 155 Photometric evaluation value generation unit 156 Luminance distribution unit 157 Boundary luminance value generation unit 160 Camera operation parameter 170 Reference value setting unit 180 Difference value Calculation unit 190 Control unit 191 Bright side luminance group memory 192 Dark side luminance group memory 200 Computer system 210 Computer 211 Processor 212 Memory 213 Display controller 214 Input / output interface 220 Display 230 Mouse 240 Keyboard 250 Recording medium access device 260 Printer 310 Conversion unit 311 to 313 Nonlinear conversion unit 320 Gradation correction processing unit 321 to 323 Gradation correction unit 330 Inverse conversion unit 331 to 333 Nonlinear inverse Conversion unit 361 Luminance region generation unit 362 Spline generation unit 363 Tone curve memory 364 γ_comp parameter memory 371 Luminance generation unit 372 Non-linear conversion unit 377 Interpolation unit 380 Tone compression unit 381, 382 Mapping unit 390 Contrast correction unit

Claims (5)

For each partial region in the input image signal composed of a plurality of pixels, a contrast based on the input luminance that is the luminance of the input image signal is calculated as a local contrast, and a partial region in which the local contrast is greater than a predetermined value is detected as a contrast region. Contrast region detection means;
Luminance distribution means for distributing the bright luminance of the input luminance to the bright dark side luminance of the input luminance and the dark dark luminance of the input luminance in the contrast region;
Boundary luminance value generation means for generating a boundary luminance value of the bright side luminance and the dark side luminance based on the distribution of the bright side luminance and the dark side luminance;
Brightness correction means for performing brightness correction so that the input brightness value corresponding to the boundary brightness value is output as a moderate brightness value;
The brightness correction means includes
Conversion curve calculation means for calculating a conversion curve used for compression of luminance gradation based on the boundary luminance value and the distribution of the input luminance;
A global luminance calculating means for calculating a global luminance which is a luminance of a global luminance image composed of low frequency components of the input image;
A gradation compression means for compressing the gradation of the input luminance and the gradation of the overall luminance based on the conversion curve;
An image output apparatus comprising: a contrast correction unit configured to correct a contrast of a gradation-compressed input image including the input luminance whose gradation has been compressed based on the global luminance whose inclination and gradation are compressed. The boundary luminance value generating means includes an average value E _{G = 1} of the bright side luminance, a variance value V _{G = 1} of the bright side luminance, an average value E _{G = 0 of the} dark side luminance, and the dark side Equation based on luminance variance VG _{= 0}

The image output apparatus according to claim 1, wherein a boundary luminance value I _m (E _{G = 0} <I _m <E _{G = 1} ), which is a solution of the above, is generated as the boundary luminance value. The boundary luminance value generating means includes an average value E _{G = 1} of the bright side luminance, a variance value V _{G = 1} of the bright side luminance, an average value E _{G = 0 of the} dark side luminance, and the dark side Equation based on luminance variance VG _{= 0}

The image output apparatus according to claim 1, wherein a boundary luminance value I _m (E _{G = 0} <I _m <E _{G = 1} ), which is a solution of the above, is generated as the boundary luminance value. For each partial region in the input image signal composed of a plurality of pixels, a contrast based on the input luminance that is the luminance of the input image signal is calculated as a local contrast, and a partial region having the local contrast greater than a predetermined value is detected as a contrast region. Contrast area detection procedure;
A luminance distribution procedure for distributing the bright brightness of the input brightness to the dark dark brightness of the input brightness in the contrast region;
A boundary luminance value generation procedure for generating a boundary luminance value of the bright side luminance and the dark side luminance based on the distribution of the bright side luminance and the dark side luminance;
A brightness correction procedure for performing brightness correction so that the input brightness value corresponding to the boundary brightness value is output as a moderate brightness value,
The brightness correction procedure is as follows:
A conversion curve calculation procedure for calculating a conversion curve used for compression of luminance gradation based on the boundary luminance value and the distribution of the input luminance;
A general luminance calculation procedure for calculating a general luminance which is a luminance of a general luminance image composed of low frequency components of the input image;
A gradation compression procedure for compressing the gradation of the input luminance and the gradation of the overall luminance based on the conversion curve;
An image output method comprising: a contrast correction procedure for correcting a contrast of a gradation-compressed input image composed of the input luminance with the gradation compressed based on the global luminance with the gradient of the conversion curve and the gradation compressed. For each partial region in the input image signal composed of a plurality of pixels, a contrast based on the input luminance that is the luminance of the input image signal is calculated as a local contrast, and a partial region having the local contrast greater than a predetermined value is detected as a contrast region. Contrast area detection procedure;
A luminance distribution procedure for distributing the bright brightness of the input brightness to the dark dark brightness of the input brightness in the contrast region;
A boundary luminance value generation procedure for generating a boundary luminance value of the bright side luminance and the dark side luminance based on the distribution of the bright side luminance and the dark side luminance;
Causing the computer to execute a brightness correction procedure for performing brightness correction so that the input brightness value corresponding to the boundary brightness value is output as a moderate brightness value;
In the brightness correction procedure,
A conversion curve calculation procedure for calculating a conversion curve used for compression of luminance gradation based on the boundary luminance value and the distribution of the input luminance;
A general luminance calculation procedure for calculating a general luminance which is a luminance of a general luminance image composed of low frequency components of the input image;
A gradation compression procedure for compressing the gradation of the input luminance and the gradation of the overall luminance based on the conversion curve;
A program for causing a computer to execute a contrast correction procedure for correcting the contrast of a gradation-compressed input image composed of the input luminance whose gradation is compressed based on the global luminance in which the gradient and gradation of the conversion curve are compressed.

Images