JP2006128885A - Image correction device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To process image without spoiling it by reducing an influence of correction of each correction area upon other correction areas when setting a plurality of correction areas one over another. <P>SOLUTION: In an image correction device, vector length representative of a magnitude of a vector which first and second color difference signal components on a color difference plane of each input pixel should form is calculated, a luminance signal component and the vector length of the input pixel are used to detect the degree of density of color signal components of the input pixel, and the density output and a first correction gain are multiplied, and a preliminarily set external parameter is subtracted from the product to obtain a first correction value, and the first correction value is subtracted from one to obtain a second correction value, and the first correction gain is adjusted by the first correction value, and a second correction gain is adjusted by the second correction value. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image correction apparatus.

  There are various types of displays for displaying video signals, and the chromaticity points of the three primary colors vary between the displays. Further, even for the same type of display, chromaticity points vary among individual displays. Further, even in the video signal generated by the imaging device, a shift in color tone (hue or saturation) or gradation (luminance) occurs due to variations in the three primary colors of the imaging device and differences in imaging conditions. Therefore, correction of hue, saturation, and luminance may be necessary.

  By the way, the baseband video signal includes three primary color signals of red (R), green (G), and blue (B), a luminance signal (Y), and two color difference signals (RY, BY, or Pb, Or Pr) is used. There are three primary color signals R, G, and B to be displayed on the display. However, when the hue and saturation are corrected, the color difference signal is easier to process than the three primary color signals. Generally, a color difference signal is used as a signal for performing saturation correction processing. Note that the three primary color signals, the luminance signal, and the color difference signal can be easily converted from each other by linear matrix calculation.

  As an example of conventional hue and saturation correction processing, there are arithmetic processing shown in the following equations (1) and (2). In equations (1) and (2), RY and BY are color difference signals before correction, ry and by are color difference signals after correction, and A1 and A2 are coefficients. As is well known, the two color difference signals are represented by a plane (color difference plane) composed of two orthogonal axes. T1 and T2 in the expressions (1) and (2) are correction angles on the color difference plane.

(by) = A1 × cos (T1) × (BY) + A1 × sin (T2) × (RY) (1)
(ry) =-A2 × sin (T1) × (BY) + A2 × cos (T2) × (RY) (2)
In the equations (1) and (2), when the coefficients A1 and A2 are different, the amplitude (saturation) of the color difference signals ry and by can be set individually. When the angles T1 and T2 are different, since the rotation angles of the color difference signals RY and BY are different, the hue can be corrected unevenly. If the coefficients A1 and A2 and the angles T1 and T2 are the same, saturation and hue can be corrected uniformly. In any case, in the correction processing according to the equations (1) and (2), the entire area on the color difference plane with the two color difference signals as axes changes.

  As other color correction circuits for correcting hues and saturations in a specific range, which are other conventional examples, disclosed in JP-A-10-145805 (Patent Document 1) and JP-A-2001-128189 (Patent Document 2). Is known.

  In the conventional correction processing using the equations (1) and (2), the entire hue is corrected, and therefore the hue and saturation cannot be corrected only in a specific angle area on the color difference plane, which is a specific hue area. Although there was a problem, the above-described Patent Document 2 discloses a correction process for avoiding this problem. According to this color correction process, only a specific angle region can be set as a correction region, and the hue and saturation of the pixels in the correction region can be corrected. However, in this color correction process, the hues of all the pixels in the angle region surrounded by the two equal hue lines are corrected in a direction perpendicular to the equal hue line at the center of the two equal hue lines, The saturation of all the pixels is corrected in a direction parallel to the central equi-hue line. For this reason, when the angle area as the correction area becomes large, the hue and saturation are corrected correctly in the vicinity of the central equi-hue line, but the hue and saturation are sufficiently corrected in the area away from the central equi-hue line. However, improvement is still necessary to make a wide angle region a correction region.

  In the conventional example of Patent Document 2, although the hue and saturation can be corrected, there is a technical problem that the luminance signal (gradation) cannot be corrected together with the correction of the hue and saturation. Furthermore, when the luminance signal is corrected, the saturation is apparently changed. Therefore, the saturation should be corrected together with the correction of the luminance signal. However, the saturation changing with the correction of the luminance signal cannot be corrected. There were also technical challenges.

  In view of such a technical problem, the specification of Japanese Patent Application No. 2003-319482, which is filed by the applicant of the present application, correctly corrects hue and saturation even when the correction area is a wide angle area. A video correction device that can be used has been proposed. According to this, the correction gain adjusting means can weaken the correction gain of the portion close to the skin color in the red region and the yellow region, and can reduce the influence of the correction of each color in the red region and the yellow region on the skin color.

However, in this proposed video correction device, in particular, when the skin color and yellow or the skin color and red correction regions overlap, whether or not to perform the skin color correction is determined by whether the input video signal is in the skin color correction region. The difference between the correction gain for correcting the skin color area and the correction gain for correcting the red or yellow area is made by using the binary value of whether or not the image is present and the binary value of whether the color density is larger or smaller than a predetermined fixed value. When the image is large, the difference between the correction areas is large, which causes a color difference when the screen is viewed with an actual video signal, and the technical problem that the image may look like noise as a whole remains. It had been. In addition, when the input signal has low brightness and low saturation, the correction range and effects are larger than those for medium and high brightness and medium and high saturation, and the correction results differ depending on the correction area of the input signal. In other words, there is still a technical problem that color steps appear on the screen, which may appear like noise as a whole.
JP-A-10-145805 JP 2001-128189 A

  The present invention has been made in view of the technical problems as described above, and even when a plurality of correction areas are set in an overlapping manner, the effect of the correction in each correction area on the other correction areas is reduced. It is an object of the present invention to provide an image correction apparatus that can perform processing without causing a failure.

  According to the first aspect of the present invention, one of the first and second color difference signals, which are color signal components of the pixels of the video signal, is used as the first axis, and the other of the first and second color difference signals is used as the first axis. At least first and second angle regions on the color difference plane formed by the first and second axes are set as the first and second correction regions, and the first axis And an angle calculation for calculating an angle formed by the first color difference signal component and the second color difference signal component on the color difference plane of each input pixel in the video correction device for correcting the respective video signals in the second correction region. And a correction function using the angle of each input pixel calculated by the angle calculation unit as a parameter, and at least one of hue, saturation and luminance of the pixel of the video signal is corrected outside the correction region. The correction gain for A correction gain for correcting at least one of hue, saturation, and luminance of a pixel in the correction area is calculated based on a correction gain calculation formula that generates a predetermined correction gain in the correction area. First and second correction gain calculation means provided for each of the first and second correction regions, the first color difference signal component and the second color difference signal on the color difference plane of each of the input pixels. A vector length calculating means for calculating a vector length representing the magnitude of a vector formed by the component, a luminance signal component of each of the input pixels and the vector length, and a degree of shading of the color signal component of each of the input pixels Gray level detection means for detecting the output, first multiplication means for multiplying the output of the gray level detection means by the first correction gain output from the first correction gain calculation means, and the output of the first multiplication means Product Subtracting a preset external parameter from the first subtraction means to obtain a first correction value, and subtracting the first correction value output from the first subtraction means from 1, the second correction Second subtracting means for obtaining a second correction value for the area, and first correction gain adjusting means for adjusting the correction gain of the first correction area by the first correction value output from the first subtracting means; And a second correction gain adjusting means for adjusting the correction gain of the second correction area according to the second correction value output from the second subtracting means.

  According to a second aspect of the present invention, there is provided the video correction device according to the first aspect, wherein a selector for setting a correction range threshold value in low saturation by an external parameter preset for the output of the vector length calculation means; and the selector Second multiplication means for multiplying the output of the first correction gain by the first correction gain, third multiplication means for multiplying the output of the selector by the second correction gain, and first output of the first subtraction means. And a product value output from the second multiplication means to obtain a new first correction value, which is supplied to the first correction gain adjustment means, and a second multiplication means. The second correction value output from the subtracting means is multiplied by the product value output from the third multiplying means to obtain a new second correction value, which is supplied to the second correction gain adjusting means. The multiplication means is provided.

  According to the video correction apparatus of the present invention, even when a plurality of correction areas are set in an overlapping manner, the influence of correction in each correction area on other correction areas can be reduced. Further, according to the present invention, it is possible to reduce the influence of correction at low luminance and low saturation on the entire screen.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the video correction apparatus shown in FIG. 1, color difference signals RY and BY are input to the input terminals 1 and 2, respectively, and the color difference signals RY and BY are input to the angle calculation unit 4. A luminance signal Y is input to the input terminal 3, and this luminance signal Y is input to the luminance / color difference gamma processing unit 19. The color difference signals RY and BY and the luminance signal Y are sequentially input for each pixel of the video signal to be corrected. When the video signal to be corrected is a three-primary color signal, the angle calculation unit converts the three primary color signals into color difference signals RY, BY and luminance signal Y in advance by linear matrix calculation. 4 and the luminance / color difference gamma processing unit 19.

  As shown in FIG. 3, one of the color difference signals in the color difference signals RY and BY (for example, BY) is the horizontal axis, and the other color difference signal (for example, RY) is orthogonal to the horizontal axis. In this case, the color difference signals RY and BY are represented by a color difference plane. The rotation direction around the intersection of the RY axis and the BY axis of this color difference plane represents the hue, and the radial direction represents the saturation. The color signal component of each pixel of the video signal is represented by color difference signals RY and BY, and each pixel is located on one of the color difference planes. Further, colors such as red, yellow, and green are located on any color difference plane, and each color can be represented by an angle θ from the BY axis.

  The angle calculation unit 4 calculates the angle formed on the color difference plane of FIG. 3 for each input pixel by a process as shown in FIG. 2 as an example. In FIG. 2, the angle calculation process in the angle calculation unit 4 is illustrated in a flowchart, but the angle calculation process in the angle calculation unit 4 may be realized by either software or hardware. In FIG. 2, in step S401, it is detected in which quadrant on the color difference plane it is based on the sign of the color difference signals RY and BY components of the input pixels. In step S402, the larger one of the absolute values of the color difference signals RY and BY components is calculated as A, and the smaller one is calculated as B.

In step S403, the angle T1 is detected from B / A. This angle T1 is 0 to 45 °, as is apparent from the processing in step S402. The angle T1 can be calculated by broken line approximation or a ROM table. In step S404, it is determined whether A is | R−Y |, that is, whether | R−Y |> | B || If | R−Y |> | B−Y | is not satisfied, the angle T1 becomes the angle T as it is. If | R−Y |> | B−Y |, T = 90−T1 is calculated in step S405. Accordingly, tan ⁻¹ ((R−Y) / (B−Y)) is obtained. The angle T1 detected in step S403 is set to 0 to 45 ° because the tan ⁻¹ ((R−Y) / (BY)) curve suddenly increases when the curve exceeds 45 °. This is because it is unsuitable for calculation.

  Further, in step S406, it is determined whether or not the quadrant is the second quadrant using the quadrant data detected in step S401. If the quadrant is the second quadrant, T = 180−T1 is calculated in step S407. If it is not the second quadrant, it is determined in step S408 whether or not it is the third quadrant. If it is the third quadrant, T = 180 + T1 is calculated in step S402. If it is not the third quadrant, it is determined in step S410 whether or not it is the fourth quadrant. If it is the fourth quadrant, T = 360−T1 is calculated in step S411. Finally, in step S412, the angle T formed on the color difference plane of FIG. 3 for each input pixel is output.

  Through the above processing, the angle of the input color difference signals RY and BY on the color difference plane can be obtained in the range of 0 to 360 °. Steps S404 to S411 are processes for correcting the angle T1 detected in step S403 to the angle T. In steps S404 to S411, it can be seen that the angle T1 is corrected according to the first to fourth quadrants. At this time, in the first quadrant, the correction processing from the angle T1 to the angle T differs depending on whether or not | R−Y |> | B−Y |.

  In the angle calculation process by the angle calculation unit 4 described above, the angle formed by the color difference signals RY and BY is calculated as an angle T1 in the range of 0 to 45 °, and the angle T1 is calculated according to each quadrant. After correction, the angle formed by the color difference signals RY and BY is calculated as T in the range of 0 to 360 °.

  When the input video signal includes a lot of noise, the color difference signals RY and BY used for angle calculation, that is, the color difference signals RY and BY input to the angle calculation unit 4 are used. It is preferable to apply a low-pass filter to reduce noise. In this way, the angles of the color difference signals RY and BY can be correctly calculated.

  1, the angle T output from the angle calculation unit 4 is input to the correction gain calculation units 51, 52,..., 5n. Here, n correction gain calculation units are provided, but one or any plurality of correction gain calculation units are provided. This correction gain calculation unit is provided in accordance with the number of regions for correcting hue and saturation. Therefore, in the case of correcting only a specific color that is one angle region on the color difference plane, only one correction gain calculation unit is required. All of the correction gain calculation units 51 to 5n have the same configuration.

Here, functions of the correction gain calculation units 51 to 5n will be described. In FIG. 3, L0, L1, and L2 are equal hue lines extending from the intersection of the BY axis and the RY axis, and the angles of the equal hue lines L0, L1, and L2 are θ ₀ , θ ₁ , θ, respectively. ₂ . The equi-hue line L0 is a correction center line, and a range of angles θ _{1 to} θ ₂ surrounded by the equi-hue lines L1 and L2 is a correction region for correcting the hue, saturation, or luminance. θ _{0 to} θ ₂ can be arbitrarily set. However, it is (theta) ₁ <(theta) ₀ <(theta) ₂ .

The angle formed by the uniform hue lines L0 and L1 (θ ₀ −θ ₁ ) and the angle formed by the uniform hue lines L0 and L2 (θ ₂ −θ ₀ ) may not be the same, but are preferably the same, etc. when the angle of hue lines L0, angle and equal hue line L0 Ll forms, L2 forms the same, set the correct angle R from the angle theta _0, the _{θ 0 -R θ 1, θ 0} + R May be θ ₂ . It should be noted that the same hue lines similar to the uniform hue lines L0 to L2 may be set at a plurality of angular positions on the stage where a plurality of correction regions are provided on the color difference plane.

Angles θ ₀ , θ ₁ , θ ₂ , or angles θ ₀ and R that determine the respective correction regions are input to the correction gain calculators 51 to 5 n. In FIG. 1, the units supplied to the correction gain calculation units 51 to 5n are θ ₀ , θ ₁ , θ ₂ and their angles θ ₀ , θ ₁ , θ ₂ are θ ₀₁ , θ ₁₁ , θ ₂₁ -θ _0n , They are distinguished from θ _1n and θ _2n . The angles θ _{01 to} θ _0n are collectively referred to as an angle θ ₀ , the angles θ _{11 to} θ _1n are collectively referred to as an angle θ ₁ , and the angles θ _{21 to} θ _2n are collectively referred to as an angle θ ₂ . The correction area can be set in an overlapping manner.

The function of the correction gain calculation units 51 to 5n will be further described with reference to FIG. The function indicating a large part of the angle than isochromatic line L1 in FIG. 3 and F _L1, which is represented by C language,
F _L1 = T-θ ₁ ;
if (F _L1 <0) F _L1 = 0;… (3)
It becomes. This equation (3) has the characteristics shown in FIG.

Similarly, a function indicating a portion having a smaller angle than the constant hue line L2 in FIG. 3 is denoted as _FL2, and this is expressed in C language.
F _L2 = θ ₂ -T;
if (F _L2 <0) F _L2 = 0;… (4)
It becomes. This equation (4) has the characteristics shown in FIG.

  The correction gain calculators 51 to 5n generate a first correction function M12 for correcting the video (hue, saturation, luminance) based on the following equation (5).

M ₁₂ = Min (F _L1 , F _L2 ); (5)
The first correction function M12 is (3) is intended to select the smaller of the F _L2 indicated by the F _L1 (4) expression of formula, the angle theta ₀ as shown in FIG. 4 (C) It has a triangular characteristic with a vertex at. The characteristic of the first correction function M12 is not limited to a triangular shape, and may be trapezoidal by providing an upper limit value or may be a cosine function. The first correction function M12 needs to set the outside of the angles θ ₁ and θ ₂ outside the correction range to 0.

Similarly, a function indicating a portion having a larger angle and a portion having a smaller angle than the constant hue line L0 in FIG. 3 is represented by _FL0, and this is expressed in C language.
F _L0 = θ ₀ -T; (6)
It becomes. This equation (6) has the characteristics shown in FIG.

  The correction gain calculation units 51 to 5n generate a second correction function M012 for correcting the video (hue, saturation, luminance) based on the following equation (7).

M012 = Min (F _L0 , F _L1 );
M012 = Max (M012, -F _L2 ); (7)
The second correction function M012 is (6) to select the smaller of the _{F L1} indicated by _{F L0} and (3) represented by the formula further, larger the M012 and _{-F L2} obtained in this Is to select. As shown in FIG. 4E, the second correction function M012 obtained in this way has an upwardly convex triangular shape having a positive vertex at the center of the angle θ ₁ and the angle θ ₀ , and the angle The characteristic is a combination of a downwardly convex triangular shape having a negative apex at the center of θ ₀ and angle θ ₂ . The second correction function M012 also needs to set the outside of the angles θ ₁ and θ ₂ outside the correction range to 0.

By the way, when θ ₁ = θ ₀ −R and θ ₂ = θ ₀ + R, the angle T calculated by the angle calculator 4 is a value of 0 to 360 °, and the angles θ ₁ and θ ₂ are 0 °. Or when it crosses 360 degrees, a discontinuity arises in the value of an angle. Therefore, the correction gain calculation units 51 to 5n correct the angle T used in the above expressions (3), (4), and (6) in advance by the following expression (8) expressed in C language. In the formula (8),> = means ≧.

if (T-θ ₁ > = 360) T = T-360;
if (θ ₂ −T> = 360) T = T + 360;… (8)
According to the equation (8), even when the angle θ ₁ and the angle θ ₂ cross over 0 ° or 360 °, when the angle T is included in the range of the angles θ _{1 to} θ ₂ , the angle T is continuously set. is corrected to a value, it is possible to make discrete value of the angle T in the range of the angle theta ₁ through? ₂ does not occur. Therefore, the first and second correction functions M12 and M012 shown in FIGS. 4C and 4E can be correctly generated.

As described above, in the present embodiment, when θ ₁ = θ ₀ −R and θ ₂ = θ ₀ + R are set as the correction regions centered on the angle θ ₀ , the correction angle R is increased to about 180 °. That is, a correction area of θ ₀ ± 180 ° can be set.

In addition to the angles θ ₀ , θ ₁ , and θ ₂ , coefficients p 1, p 2, s 1, y 1, and cl are input to the correction gain calculators 51 to 5 n. The coefficients p1 and p2 are coefficients for changing the hue correction gain described later, and the coefficient s1 is a coefficient for changing the saturation correction gain described later. The coefficient yl is a coefficient for changing the luminance gamma correction gain described later, and the coefficient c1 is a coefficient for changing the saturation gamma correction gain described later. The coefficients p1, p2, s1, y1, and c1 input to the correction gain calculation units 51 to 5n do not have to be the same as those of the correction gain calculation units 51 to 5n. When the coefficients p1, p2, s1, y1, and c1 are individually set for the correction gain calculation units 51 to 5n, the degree of image correction in the plurality of correction regions can be varied.

  In the present embodiment, not only the hue and saturation are corrected, but also the luminance signal is corrected. Further, when the luminance signal is corrected, the saturation is apparently changed. Therefore, the saturation is corrected together with the correction of the luminance signal. As will be described later, in the present embodiment, the correction of the luminance signal is to give the luminance signal a gamma curve characteristic. In addition, the saturation correction combined with the correction of the luminance signal also gives the color difference signal the same gamma curve characteristic as the luminance correction. Therefore, in the present embodiment, the correction of the luminance signal is referred to as Y gamma, and the saturation correction combined with the correction of the luminance signal is referred to as C gamma.

  The correction gain calculators 51 to 5n use the first and second correction functions M12 and M012 generated as described above and the input coefficients p1, p2, s1, y1, and c1, respectively, Correction gains P, S, Gy, and Gc are generated based on correction gain calculation expressions P (T), S (T), Gy (T), and Gc (T) shown in Equations 9) to (12). P is a hue correction gain used to correct the hue gain, S is a saturation correction gain used to correct the saturation gain, Gy is a Y gamma correction gain used to correct the Y gamma gain, Gc Is a C gamma correction gain used to correct the C gamma gain.

P (T) = pl × M012 + p2 × M12; (9)
S (T) = sl × M12; (10)
Gy (T) = y1 × M12; (11)
Gc (T) = cl × M12; (12)
As described with reference to FIG. 4, since the first and second correction functions M12 and M012 are functions having the angle T calculated by the angle calculation unit 4 as a parameter, the equations (9) to (12) are used. The correction gain calculation formula shown below is a function with the angle T as a parameter. Figure 4 first correction function M12 shown in (C) has a central angle theta ₀ and is equal hue line closer to L0 correction amount increases characteristic of the correction region, the pixels in the correction area in one direction It has the characteristic to correct.

  On the other hand, in the second correction function M012 shown in FIG. 4E, the correction amount is 0 on the equal hue line L0, and the correction amount is maximum at the center of each of the equal hue line L0 and the equal hue lines L1 and L2. It has the characteristic which becomes. In addition, in the angle region sandwiched between the equal hue line L0 and the equal hue line L1, and in the angle region sandwiched between the equal hue line L0 and the equal hue line L2, the correction directions have reverse characteristics. The coefficients p1, p2, s1, y1, c1 vary the amplitudes of the first and second correction functions M12, M012, and the correction gain calculation formulas P (T), S (T), Gy (T), Gc It can be seen that the correction gains P, S, Gy, and Gc generated by (T) are set appropriately.

  Furthermore, the calculation of the correction gain by the hue correction gain calculation formula P (T) of the equation (9) will be described. As shown in FIG. 5A, pl × M012 which is the first term of the equation (9) shifts the hue in the correction area toward the equi-hue line L0 (or in the opposite direction). . This is apparent from the characteristics of the second correction function M012. At this time, for the reason described later, the chromaticity of each pixel in the angle region sandwiched between the equi-hue line L0 and the equi-hue line L1 and in the angle region sandwiched between the equi-hue line L0 and the equi-hue line L2. The points are substantially rotated in opposite directions.

  Further, p2 × M12, which is the second term of the equation (9), shifts the hue in the correction region in one direction as shown in FIG. 5B. As is clear from the characteristic of the first correction function M12, the shift amount increases as it approaches the constant hue line L0. At this time, for the reason described later, the chromaticity of each pixel in the angle region sandwiched between the equal hue line L0 and the equal hue line L1 and in the angle region sandwiched between the equal hue line L0 and the equal hue line L2. The points are rotated in substantially the same direction as each other.

The hue correction according to the equation (9) is performed only within the correction region surrounded by the angles θ _{1 to} θ ₂ , and the influence of the correction does not reach at all outside the correction region.

  In this embodiment, the first and second correction functions M12 and M012 are generated, and the hue correction gain P is generated by the correction gain calculation formula P (T) of the formula (9). By providing two functions in FIG. 5B adjacent to each other on the color difference flat image and shifting the hues in opposite directions, the function becomes equivalent to the function in FIG. Therefore, a correction gain calculation expression equivalent to the expression (9) can be obtained by using only the first correction function M12 without generating the second correction function M012. In the case of such a configuration, the generation of the second correction function M012 becomes unnecessary.

As shown in the above equations (10) to (12), in the correction gain calculation formulas S (T), Gy (T), and Gc (T) for calculating the correction gains of saturation, Y gamma, and C gamma, The first correction function M12 is used. Correction of the saturation, Y gamma, and C gamma according to the equations (10) to (12) is also performed only in the correction area surrounded by the angles θ _{1 to} θ ₂ , and the influence of the correction is completely exerted outside the correction area. Absent.

  In FIG. 1, the output of the correction gain calculation units 51 to 5n and the hue, saturation, and the like generated by the correction gain calculation formulas P (T) to Gc (T) shown in the above formulas (9) to (12). The correction gains for Y gamma and C gamma are Pi, Si, Gyi, and Gci (i = 1 to n), respectively. In FIG. 1, Pi, Si, Gyi, and Gci are simply described as P, S, Gy, and Gc. The correction gains Si, Pi, Gyi, and Gci output from the correction gain calculators 51 to 5n are accumulated (summed) by accumulators 6 to 9, respectively. Assume that the outputs of the accumulators 6 to 9 are ΣSi, ΣPi, ΣGyi, and ΣGci, respectively.

  If the chromaticity point of the input pixel is included in any one of the correction areas, the correction gains Si, Pi, Gyi with respect to the color difference signals RY, BY and the luminance signal Y are calculated by a subsequent calculation process. , Gci is used for correction. In the present embodiment, even when there are a plurality of correction gain calculation units, that is, when a plurality of correction regions are set on the color difference plane, the respective correction gains Si, Pi, Gyi, Gci are accumulated. Therefore, the circuit configuration of the correction calculation processing for the subsequent color difference signals RY and BY and the luminance signal Y is simplified. If the chromaticity point of the input pixel is included in a plurality of correction areas, the correction gain is the sum, and if the chromaticity point of the input pixel is not included in any correction area, the correction gain is 0.

  The multiplier 10, adder 13, and inverter 14 receive the color difference signal RY input to the input terminal 1, and the multipliers 11, 15 and adder 18 input the color difference signal input to the input terminal 2. BY is input. The output of the inverter 14 is input to the multiplier 16. The correction gain ΣSi output from the accumulator 6 is input to the multipliers 10 and 15. The multiplier 10 multiplies the color difference signal RY by the correction gain ΣSi and outputs the result, and the multiplier 15 multiplies the color difference signal BY by the correction gain ΣSi and outputs the result. The correction gain ΣPi output from the accumulator 7 is input to the multipliers 11 and 16. The multiplier 11 multiplies the color difference signal BY by the correction gain ΣPi and outputs the result. The multiplier 16 multiplies the color difference signal RY inverted by the inverter 14 by the correction gain ΣPi and outputs the result.

  The adder 12 adds the outputs of the multipliers 10 and 11, and the adder 13 adds the color difference signal RY and the output of the adder 12. The output of the adder 13 is a color difference signal RY after correction. The adder 17 adds the outputs of the multipliers 15 and 16, and the adder 18 adds the color difference signal BY and the output of the adder 17. The output of the adder 18 is a corrected color difference signal BY.

  The multipliers 10 and 11 and the adders 12 and 13 calculate the following equation (13), and the multipliers 15 and 16, the inverter 14, and the adders 17 and 18 calculate the following equation (14). You can see that it is processing. The color difference signal RY corrected by the equation (13) is obtained, and the color difference signal BY corrected by the equation (14) is obtained.

ΣPi × (BY) + ΣSi × (RY) + (RY) (13)
-ΣPi × (RY) + ΣSi × (BY) + (BY) (14)
As described with reference to FIG. 3, the chromaticity points of the input pixels are represented by (BY) and (RY). Considering the chromaticity point as a vector from the origin of the color difference plane in FIG. 3, vectors orthogonal to the vector are-(RY) and (BY). Accordingly, −ΣPi × (R−Y) and ΣPi × (B−Y) in the first term of the equations (14) and (13) are orthogonal to the vector of chromaticity points of the pixel to be corrected. It can be seen that this represents a hue correction component in the direction of the image. In addition, ΣSi × (BY) and ΣSi × (RY) in the second term of the equations (14) and (13) are in the same direction as the chromaticity point vector of the pixel to be corrected. It can be seen that this represents a correction to, that is, a saturation correction component.

  Therefore, in the present embodiment, in hue correction, the chromaticity point of each pixel in the correction area is connected to the intersection of the RY axis and the BY axis and each pixel in the correction area. It is moved in the direction orthogonal to the hue line. Therefore, the hue substantially rotates and moves within the entire correction area. In the saturation correction, the chromaticity point of each pixel in the correction area is parallel to each equal hue line connecting the intersection of the RY axis and the BY axis and each pixel in the correction area. (Either in the direction of increasing or decreasing the saturation on the equi-hue line). Therefore, in the present embodiment, even when the angle region as the correction region is set large, it is possible to correct the hue and saturation correctly as intended.

  As described above, the first circuit block of the multipliers 10 and 11 and the adders 12 and 13, the multiplier 15 and 16, the inverter 14, and the second circuit block of the adders 17 and 18, the color difference signal. By calculating RY, B 1 Y and correction gains ΣPi, ΣSi, color difference signals RY, BY with hue and saturation corrections are generated. These first circuit block and second circuit block operate as correction means for correcting hue and saturation. The color difference signals RY and BY subjected to the hue and saturation corrections output from the adders 13 and 18 are input to the luminance / color difference gamma correction unit 19.

  By the way, the correction of the color difference signals RY and BY by the above-described equations (13) and (14) is performed by connecting the intersection of the RY axis and the BY axis and each pixel in the correction area. It is moved in the direction orthogonal to the equi-hue line. When the correction amount (movement amount) is small, it can be considered that the pixels in the correction region substantially move in the rotation direction. However, if the correction amount is large, the pixels in the correction area cannot be regarded as moving in the rotation direction. When the pixel moves greatly in the direction orthogonal to the equi-hue line, the saturation also changes.

  For this reason, when increasing the correction amount, it is preferable to move the chromaticity point of each pixel in the correction area in the rotation direction on the color difference plane. Expressions for moving the chromaticity points of the pixels in the rotation direction on the color difference plane can be expressed by the following expressions (15) and (16). The color difference signal RY corrected by the equation (15) is obtained, and the color difference signal BY corrected by the equation (16) is obtained.

(1 + ΣSi) × sin (ΣPi) × (BY) + (1 + ΣΣSi) × cos (ΣPi) × (RY) (15)
(1 + ΣSi) × cos (ΣPi) × (BY) − (1 + ΣΣSi) × sin (ΣPi) × (RY) (16)
Although the equations (15) and (16) may be used as they are, the sine function and cosine function in the equations (15) and (16) may be approximated as the following equations (17) and (18). . According to the equations (17) and (18), ΣPi approximates very well in the range up to π / 2 radians, and can be said to be practically sufficient.

sin (ΣPi) = ΣPi− (ΣPi) ³ /6.6 (17)
cos (ΣPi) = 1- (ΣPi) ² /2.2 (18)
When moving the chromaticity points of the pixels in the correction area in the rotation direction on the color difference plane, the first circuit block of the adders 10 and 11 and the adders 12 and 13, the multipliers 15 and 16, and the inverter 14 and the second circuit block portion of the adders 17 and 18 may be replaced with a circuit configuration for realizing the equations (15) and (16). Of course, the equations (15) and (16) may be realized by software.

  In FIG. 1, the luminance / color difference gamma correction unit 19 receives the luminance signal Y input to the input terminal 3 and the correction gains ΣGyi and ΣGci output from the accumulators 8 and 9 as described above. As shown in FIG. 6, the luminance / color difference gamma correction unit 19 includes a correction function represented by the following equation (19).

G (X) = X ^{1 / 1.5} -X (19)
The luminance / color difference gamma correction unit 19 performs gamma correction on the luminance signal Y and the color difference signals RY and BY, and X in the equation (19) (the horizontal axis in FIG. 6) represents the luminance signal. Y and the values of the color difference signals RY and BY. When gamma correction is performed on a video signal, optimum characteristics differ depending on the display device (display). The equation (19) is an example, and the correction characteristics in the luminance / color difference gamma correction unit 19 are not limited to the equation (19).

  The luminance / color difference gamma correction unit 19 multiplies the correction gain ΣGyi by the correction function G (X), and adds the multiplication result to the luminance signal Y. That is, the arithmetic processing of the following equation (20) is performed.

Y + G (X) × ΣGyi (20)
Here, it can be seen that the correction gain ΣGyi is a coefficient that varies the amplitude of the correction function G (X) shown in FIG.

  When the luminance is changed by such gamma correction of the luminance signal Y, the saturation is also changed, and the saturation needs to be corrected. Therefore, the luminance / color difference gamma correction unit 19 also performs gamma correction on the input color difference signals RY and BY. The luminance / color difference gamma correction unit 19 multiplies the color difference signals RY and BY by the correction gain ΣGci, the correction function G (X) and 1 / X, and the multiplication result is used as the color difference signals RY and B−. Add to each Y. In other words, the following arithmetic processings (21) and (22) are performed.

(RY) + (RY) × G (X) × ΣGci / X (21)
(BY) + (BY) × G (X) × ΣGci / X (22)
Dividing (R−Y) × G (X) × ΣGci, (B−Y) × G (X) × ΣGci by X in the equations (21) and (22) indicates the increase / decrease ratio of saturation. This is to match the luminance increase / decrease ratio. When X is a small value such as 0.1 or less, the operation may become unstable or the S / N at low luminance may be deteriorated. Therefore, it is preferable to limit the minimum value of X to 0.1 in order to improve the stability of operation and the S / N at low luminance. In this way, saturation correction (C gamma) can be performed in accordance with the gamma processing of the luminance signal.

  The luminance signal Y subjected to Y gamma output from the luminance / color difference gamma correction unit 19 is output from the output terminal 22, and the color difference signals RY and BY subjected to C gamma are output to the output terminal 20, respectively. 21 is output. When it is necessary to return the luminance signal Y and the color difference signals RY and BY to the three primary color signals, they may be converted into the three primary color signals by linear matrix calculation.

  With the above configuration, the color difference signals RY and BY and the luminance signal Y input to the input terminals 1 to 3 can correct the hue and saturation only within a preset correction region. In addition to correction, Y gamma and C gamma can be applied. In this embodiment, the case of correcting all of hue, saturation, Y gamma, and C gamma has been described. However, only one of these elements may be corrected, or any combination may be corrected. . However, it is preferable that Y gamma and C gamma be a pair.

  In the present embodiment described above, the first circuit block of the multipliers 10 and 11 and the adders 12 and 13, and the second circuit block of the multipliers 15 and 16, the inverter 14, and the adders 17 and 18 are used. The correction of hue and saturation is irrelevant to the value of the luminance signal Y. Depending on the value of the luminance signal Y, the hue and saturation correction amounts may be varied. In order to vary the correction amount of hue and saturation according to the value of the luminance signal Y, the coefficients p1, p2, and s1 input to the correction gain calculation units 51 to 5n are changed according to the value of the luminance signal Y. Good.

  FIG. 7 shows an example of characteristics when the coefficient sl is changed in accordance with the value of the luminance signal Y to vary the saturation correction amount. The horizontal axis is the luminance signal Y, and the vertical axis is the coefficient s1. This characteristic is such that the saturation is lowered at low luminance and the saturation is increased at high luminance. As a means for changing the coefficient s1, for example, a table as shown in FIG. 8 can be used. In FIG. 8, the table 31 stores a coefficient s1 having various values. A value of the luminance signal Y is input to the table 31 as an address, and the table 31 outputs a coefficient s1 according to the input value of the luminance signal Y. The same applies when the coefficients p1 and p2 are changed according to the value of the luminance signal Y.

  As described above, when the hue and saturation are changed according to the value of the luminance signal Y, the image in the correction area is three-dimensionally corrected as shown in FIG. In FIG. 9, the Y axis indicating the value of the luminance signal Y is orthogonal to the color difference plane composed of the RY axis and the BY axis. In FIG. 9, Y0, Y1, and Y2 indicate equiluminance planes. A circled number 1 shown in FIG. 9 conceptually shows a case where the hue and saturation of a pixel indicated by a white circle are corrected to a position indicated by a black circle regardless of the value of the luminance signal Y. A circled number 2 shown in FIG. 9 conceptually shows a case where the hue and saturation of a pixel indicated by a white circle are corrected to a position indicated by a black circle according to the value of the luminance signal Y.

  Further, the circled number 3 shown in FIG. 9 conceptually shows the case where Y gamma and C gamma are applied in addition to the correction of the circled number 1. The four circle numbers shown in FIG. 9 conceptually show the case where Y gamma and C gamma are applied in addition to the correction of the circle number 2.

  In FIG. 1, the configuration of the present embodiment is described as hardware, but it is needless to say that the configuration may be realized by software (contributor program). In the present invention, the configuration of the present invention may be an integrated circuit, a program including the steps of the present invention, or a recording medium on which this program is recorded. The recording medium may be any medium including an optical disk such as CD-R0M and other disks. This is the same as in the second embodiment.

  As is clear from the above description, a video correction program for causing a video correction method or a computer to execute video correction includes the following steps. First, as a first step, an angle calculation step of calculating an angle formed by the color difference signals RY and BY components on the color difference plane of each input pixel is included.

  As a second step, with the angle of each input pixel calculated in the angle calculation step as a parameter, a correction gain for correcting the tone, saturation, Y gamma, and C gamma of the pixels of the video signal outside the correction region is 0. Correction gain calculation for calculating a correction gain for correcting the hue, saturation, Y gamma, and G gamma of a pixel in the correction area based on a correction gain calculation formula for generating a predetermined correction gain in the correction area Includes steps.

  When a plurality of correction areas are set on the color difference plane, the third step includes an accumulation step of accumulating the correction gains of the plurality of correction areas. If there is only one correction area, this accumulation step is not necessary. The fourth step includes a correction step of correcting the hue and saturation of the pixels in the correction region by calculating the color difference signals RY and BY and the hue and saturation correction gains.

  Further, as a fifth step, the luminance signal Y is corrected by adding the Y gamma correction gain multiplied by a predetermined correction function to the luminance signal Y (Y gamma correction), and the color difference signals RY, B -Y multiplied by the correction gain in C gamma and its correction function is added to each of the color difference signals RY and BY to correct the color difference signals RY and BY (C gamma correction). Step).

  The first to third steps need to be performed in this order. The fourth step and the fifth step may be performed at the same timing, and either may be before or after in time.

  By the way, the present invention can be used not only for correcting variations in chromaticity points and image variations caused by an imaging apparatus, but also for the purpose of intentionally correcting a specific color. For example, the green color of lawn or tree can be changed to yellow or brown to change the seasonal feeling of the image. As described above, the present invention can arbitrarily change the hue, saturation, and luminance (gradation) of a specific color of an image, and thus can be used for various days. In the present embodiment, the color difference signals are described as RY and BY, but the color difference signals are not limited to RY and BY, and may be color difference signals Cb and Cr. Good.

  As described above, the correction areas set by the correction gain calculation units 51 to 5n in FIG. 1 may partially overlap. When correction areas are overlapped, the above-described correction of hue, saturation, Y gamma, and C gamma affects other correction areas in which correction in one correction area partially overlaps the one correction area. It will be. When the present invention is mounted on an actual product, it may be preferable to set correction regions in an overlapping manner.

  FIG. 10 shows a preferred embodiment for reducing the influence of correction in each correction area on other correction areas when a plurality of correction areas are set in an overlapping manner. The circuit configuration shown in FIG. 10 is configured to be added to the circuit configuration of FIG. As described above, the accumulators 6 to 9 in FIG. 1 accumulate the correction gains P, S, Gy, and Gc output from the correction gain calculation units 51 to 5n, respectively. When the circuit configuration shown in FIG. 10 is added to the circuit configuration shown in FIG. 1, the correction gains P, S, Gy, and Gc replaced or adjusted by the circuit shown in FIG. The The processing subsequent to the accumulators 6 to 9 is as described above.

  FIG. 11 shows an example of a color difference plane when a plurality of correction areas are set in an overlapping manner. Consider skin color, red, and yellow as correction areas. The skin color here refers to a general Japanese skin color, which is around 125 ° on the color difference plane of FIG. 11, red is around 110 °, and yellow is around 160 °. As shown in FIG. 11, it is assumed that a region surrounded by a solid line is set as a skin color region A, a region surrounded by a one-dot chain line is set as a red region B, and a region surrounded by a two-dot chain line is set as a yellow region C. . The skin color area A and the red area B partially overlap, and the skin color area A and the yellow area C partially overlap. These three areas are set using three of the correction gain calculation units 51 to 5n.

  The video correction apparatus shown in FIG. 10 receives the color difference signals RY and BY and calculates a vector length VC formed by the color difference signals RY and BY, and each of the input pixels. The luminance signal Y is a color signal using the multiplier 61 that multiplies the external parameter to change the level in seven stages, the vector length VC output from the vector length calculation unit 60, and the luminance signal Y whose level is changed as described above. A density detector 62 for detecting the degree of density of the component, a limiter 63 for limiting the output of the density detector 62, a multiplier 64 for multiplying the output of the limiter 63 and the correction function M12, and an output of the multiplier 64 A limiter 65 for adjusting the value range, a subtracter 66 for subtracting PARA1 given as an external parameter from the output of the limiter 65, and a limit for the output of the subtractor 66. The limiter 67 for obtaining the correction value (1) of the area A, and the correction value (2) of the areas B and C by subtracting the correction value (1) output from the limiter 67 from the output of the subtractor 66. Is provided.

  The video correction apparatus according to the present embodiment further includes a limiter 71 that limits the output VC of the vector length calculation unit 60, and a selector that sets a correction range threshold value at low saturation with respect to the output of the limiter 71 using the external parameter PARA2. 72, a multiplier 73 that multiplies the output of the selector 72 by the correction range M12_s of the area A, and a limiter 75 that limits the output of the multiplier 73 and outputs a correction value (3) in the low saturation of the area A. Similarly, a multiplier 74 that multiplies the output of the selector 72 by the respective correction ranges M12_x and M12_y of the regions B and C, and limits the output of the multiplier 74 to reduce the low saturation of the regions B and C. Is provided with a limiter 76 for obtaining each correction value (4).

  The video correction apparatus according to the present embodiment further multiplies the correction value (1) for the region A output from the limiter 67 by the correction value (3) for the low saturation output from the limiter 75 to correct the entire region A. A multiplier 77 for obtaining a value (5), a limiter 78 for obtaining an adjustment value of the correction gains P, S, Gy, Gc in the region A by limiting the output of the multiplier 77, and an output of the subtractor 68 A multiplier 79 for multiplying the correction value (2) of the regions B and C to be multiplied by the correction value (4) for low saturation output from the limiter 76 to obtain the entire correction value (6) of the regions B and C, and this multiplication A limiter 80 is provided that limits the output of the device 79 to obtain adjustment values of the correction gains P, S, Gy, and Gc in the regions B and C.

  The video correction apparatus according to the present embodiment further includes a multiplier 81 that multiplies the adjustment values of the correction gains P, S, Gy, and Gc in the regions B and C output from the limiter 80 by the signals in the red region and the yellow region. A divider 82 that performs 1 / limit division on the output, a multiplier 83 that multiplies the adjustment values of the correction gains P, S, Gy, and Gc in the region A output from the limiter 78 by the skin color region signal, and the output A divider 84 that performs 1 / limit division is provided.

  Next, video correction processing by the video correction apparatus having the above configuration will be described. The color difference signals RY and BY are input to the vector length calculation unit 60. The vector length calculation unit 60 calculates a vector length VC representing the magnitude of a vector formed by the color difference signals RY and BY on the color difference plane of each input pixel. The level of the luminance signal Y of each input pixel can be changed in seven stages by the external parameter PARA3 shown in the multiplier 61, so that the correction can be made within the range of the luminance level desired by the user. The vector length VC output from the vector length calculation unit 60 and the luminance signal Y in which the level described above is changed are input to the light / dark detection unit 62. The light / dark detection unit 62 detects the light / dark degree of the color signal component by calculating Y / VC. Here, the lightness (lighter) color is set so that the output of the light / dark detection unit 62 increases. The output of the light / dark detection unit 62 is input to the limiter 63. A value obtained from the output of the limiter 63 and the correction function M12_s of the above-described equations (3) to (5) is input to the multiplier 64. The output of the multiplier 64 is input to the limiter 65 to adjust the value range.

  A value PARA1 which is the center of the boundary between the specific area A and the two areas B and C in contact with the specific area A is given as an external parameter, and the value of this external parameter and the output of the limiter 65 are input to the subtractor 66. Through the process, the correction value (1) of the area A is obtained. By subtracting the correction value (1) from the limit value of the limiter 67 by the subtractor 68, the correction values (2) of the regions B and C are obtained.

  FIG. 12 is a conceptual diagram in which the region A is skin color, B is red, and C is yellow. The horizontal axis represents Y / VC representing color shading, the vertical axis represents a numerical value indicating the degree of intersection of the areas, and the limit value of the limiter 67 is arbitrarily designated. As the width of each area from the external parameter PARA1 is increased, the crossing becomes smoother and the color correction step due to the difference in area becomes difficult to occur. However, if it is too wide, there is an adverse effect that the color correction effect is weakened.

  On the other hand, the output VC of the vector length calculation unit 60 is input to the limiter 71, and the output is set by the selector 72 using the external parameter PARA 2 to set the threshold value for the correction range in low saturation. By multiplying the output of the selector 72 by the correction range M12_s of the area A by the multiplier 73 and inputting the output to the limiter 75, the correction value (3) in the low saturation of the area A is obtained. Similarly, the output of the selector is multiplied by the respective correction ranges M12_r and M12_y of the regions B and C by the multiplier 74, and the output is input to the limiter 76, so that each of the regions B and C in the low saturation is obtained. A correction value (4) is obtained.

  Next, the correction value (1) for the area A and the correction value (3) for low saturation are multiplied by the multiplier 77 to obtain the entire correction value (5) for the area A. A value obtained by limiting this value with the limiter 78 is an adjustment value for adjusting the correction gains P, S, Gy, and Gc in the region A. Also, the obtained correction value (2) for the regions B and C and the correction value (4) for low saturation are multiplied by the multiplier 79 to obtain the entire correction value (6) for the regions B and C. A value obtained by limiting this value with the limiter 80 is an adjustment value for adjusting the correction gains P, S, Gy, and Gc in the regions B and C.

  FIG. 13 shows a conceptual diagram of low saturation gradation correction. The horizontal axis is the color difference vector VC, and the vertical axis arbitrarily designates a numerical value representing the degree of intersection of the regions and the limiter 75 or 76 value. If a large value is set in each area in the vector VC direction from the external parameter PARA2, the crossing is moderated, and when the input signal has low luminance and low saturation, there is no color step due to the difference in correction effect, and the noise is felt. Can be eliminated. Since the difference in the correction effect is conspicuous for any color region at low luminance and low saturation, it is desirable to provide a similar correction device for any color region.

  By the correction gain adjustment by the video correction apparatus, the correction gains P, S, Gy, and Gc in the red region and the yellow region that are close to the skin color are weakened. Therefore, it is possible to reduce the influence of correction of each color in the red region and the yellow region on the skin color. Although a specific area and two areas in contact with the area have been described, it is particularly effective in the case of a skin-colored area A and two areas in contact therewith, red and yellow areas B and C, as shown in FIG. These areas can be set arbitrarily.

1 is a block diagram of one embodiment of the present invention. The flowchart which shows the specific process in the angle calculation part 4 in FIG. The figure which shows a color difference plane. Explanatory drawing of the correction function used with the correction gain calculation parts 51-5n in FIG. Explanatory drawing of an effect | action of the correction function shown in FIG. Explanatory drawing of the correction function used in the brightness | luminance and color difference gamma process part 19 in FIG. The figure which shows the example of a conversion characteristic in the case of varying the coefficient of a saturation correction gain according to a luminance signal. The block diagram which shows an example of the means for changing the coefficient of a saturation correction gain according to a luminance signal. Explanatory drawing of an effect | action in the case of changing the coefficient of a hue and a saturation correction gain according to a luminance signal. The block diagram of the 2nd Embodiment of this invention. The figure which shows an example of the color difference plane at the time of setting a some correction | amendment area | region on overlapping. Explanatory drawing which shows the relationship of the correction function of each area | region in the case of setting a some correction | amendment area | region overlappingly. The conceptual diagram of low saturation gradation correction.

Explanation of symbols

60 Vector length calculation unit 61, 64, 73, 74, 77, 79 Multiplier 62 Shading detection unit 63, 65, 67, 71, 75, 76, 78, 80 Limiter 70 Low saturation gradation correction unit 81, 83 Multiplication (Correction gain adjustment means)
82,84 Divider (correction gain adjustment means)

Claims (2)

One of the first and second color difference signals, which are the color signal components of the pixels of the video signal, is a first axis, and the other of the first and second color difference signals is a second axis orthogonal to the first axis. In this case, at least the first and second angle regions on the color difference plane formed by the first and second axes are set as the first and second correction regions, and are within the first and second correction regions. In a video correction device that corrects each video signal,
Angle calculating means for calculating an angle formed by the first color difference signal component and the second color difference signal component on the color difference plane of each input pixel;
Correction for correcting at least one of hue, saturation and luminance of the pixel of the video signal outside the correction region using a correction function using the angle of each input pixel calculated by the angle calculation means as a parameter Correction for correcting at least one of hue, saturation, and luminance of a pixel in the correction area based on a correction gain calculation formula that sets the gain to 0 and generates a predetermined correction gain in the correction area First and second correction gain calculation means provided for each of the first and second correction regions for calculating the gain;
Vector length calculating means for calculating a vector length representing a magnitude of a vector formed by the first color difference signal component and the second color difference signal component on the color difference plane of each of the input pixels;
A light / dark detection means for detecting a light / dark degree of the color signal component of each input pixel using the luminance signal component of each of the input pixels and the vector length;
First multiplication means for multiplying the output of the light and shade detection means by the first correction gain output from the first correction gain calculation means;
First subtraction means for subtracting a preset external parameter from the product value output from the first multiplication means to obtain a first correction value;
Second subtracting means for obtaining a second correction value for the second correction area by subtracting the first correction value output from the first subtracting means from 1;
First correction gain adjusting means for adjusting the correction gain of the first correction area according to the first correction value output from the first subtracting means;
An image correction apparatus comprising: a second correction gain adjusting unit that adjusts the correction gain of the second correction region by the second correction value output from the second subtracting unit. A selector that sets a threshold value of a correction range in low saturation by an external parameter set in advance with respect to the output of the vector length calculation means;
Second multiplying means for multiplying the output of the selector by the first correction gain;
Third multiplication means for multiplying the output of the selector by a second correction gain;
The first correction value output from the first subtracting means is multiplied by the product value output from the second multiplying means to obtain a new first correction value, and the first correction gain adjusting means A fourth multiplying means for giving,
The second correction value output from the second subtracting means is multiplied by the product value output from the third multiplying means to obtain a new second correction value, and the second correction gain adjusting means The image correcting apparatus according to claim 1, further comprising a fifth multiplying unit that gives the image.

Images