JP2009239608A - Image processing apparatus and digital camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology for suppressing color noise contained in an image signal without exerting adverse influences upon original image information. <P>SOLUTION: A saturation evaluation circuit 20 inputs RGB pixel signals of the Bayer array output from an A/D converter 13 and calculates color strength evaluation values Rda, Gda, Bda based on pixel mean values of color component signals. The saturation evaluation circuit 20 determines a color component of a peak strength from among RGBCMY color components based on a magnitude relation of the color strength evaluation values Rda, Gda, Bda and calculates a saturation evaluation value CV by weighting the color component. The saturation evaluation value CV is converted into suppression coefficients Kb, Kr at LUTs 105, 106 and then multiply color difference signals Cb, Cr at multipliers 107, 108. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for suppressing color noise included in a captured image or the like.

  With an increase in the total number of pixels of an image sensor mounted on a digital camera, a digital movie, etc., the light receiving area of the image sensor has been miniaturized. As the light receiving area of the image sensor decreases, the light receiving sensitivity of the image sensor decreases, and as a result, the S / N ratio of the image pickup pixel signal decreases.

  In a captured image of a digital camera, a particular problem is color noise that appears to have a color such as yellow, cyan, and magenta in a portion that is originally a gray region.

  In order to suppress the color noise, a method of applying a low-pass filter or a median filter to the color difference component is widely known.

  Alternatively, in order to suppress color noise, a coring process that forcibly converts a region having a small color difference component into a gray region is used.

JP 2003-235050 A

  As described above, color noise can be suppressed by using a low-pass filter or a median filter. However, if the low-pass filter is applied too strongly, the original high-frequency component contained in the image is removed, and the sharpness of the image may be reduced. Conversely, if the low-pass filter is weakened, noise remains.

  Further, when the median filter is applied to the pixel signal, there is a side effect that the color is blurred. In other words, the color noise is reduced for the pixels in which the color noise was originally generated, but the surrounding pixels have a problem that the color noise is blurred although the signal level is low.

  Further, since the coring process is a process for forcibly converting to a gray area, there is a side effect that color loss occurs in an area where light colors originally exist.

  In Patent Document 1, the degree of adjustment of color reproducibility is controlled according to color information. Specifically, saturation enhancement is adjusted based on the saturation of the main range of the image and the average value of the saturation of the entire image. This technique adjusts the degree of saturation enhancement based on the main part of the image and the overall information.

  In view of the above problems, an object of the present invention is to provide a technique for suppressing color noise included in an image signal without adversely affecting original image information.

  In order to solve the above-mentioned problems, an invention according to claim 1 is an invention relating to an image processing apparatus, wherein an input means for inputting a pixel signal of a first color space having one color component for each pixel, and said input A saturation evaluation means for obtaining a saturation evaluation value of a pixel signal in a peripheral region including the target pixel input by the means; and interpolating the pixel signal input by the input means; for all colors in the predetermined color space for one pixel Interpolating means for generating a pixel signal having components, and a color for converting the pixel signal in the first color space interpolated by the interpolation means into a pixel signal in a second color space including a luminance signal and a color difference signal By comparing the saturation evaluation value with a predetermined condition by a space conversion means, it is determined whether or not the surrounding area is a low saturation area, and the surrounding area is determined to be a low saturation area. The color difference signal And suppressing means for suppressing issue value, characterized in that it comprises a.

  According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the suppression unit includes a multiplying unit that multiplies the color difference signal by a suppression coefficient, and a correspondence relationship between the saturation evaluation value and the suppression coefficient. And coefficient determination means for determining the suppression coefficient from the saturation evaluation value.

  According to a third aspect of the present invention, in the image processing apparatus according to the second aspect, the color difference signal includes a first color difference signal and a second color difference signal, and the coefficient determination unit is configured to output the saturation evaluation value. And a first suppression coefficient corresponding to the first color difference signal, first coefficient determination means for determining the first suppression coefficient from the saturation evaluation value, the saturation evaluation value, Second coefficient determining means for maintaining a correspondence relationship with a second suppression coefficient corresponding to a second color difference signal and determining the second suppression coefficient from the saturation evaluation value, wherein the multiplying means includes: And a first multiplying unit for multiplying the first color difference signal by the first suppression coefficient, and a second multiplying unit for multiplying the second color difference signal by the second suppression coefficient.

  According to a fourth aspect of the present invention, in the image processing apparatus according to any one of the first to third aspects, the saturation evaluation unit calculates a color intensity of each color component in the first color space. Color intensity calculation means, color component determination means for comparing each color intensity, determining one color component having the highest color intensity based on the degree of bias of each color intensity, and based on the color intensity of the one color component Evaluation value calculation means for calculating a saturation evaluation value.

  According to a fifth aspect of the present invention, in the image processing apparatus according to the fourth aspect, the color component determining unit compares the color intensities of the color components of the first color space calculated by the color intensity calculating unit. And means for determining the one color component from among the color components included in the first color space or the color components included in the color space having a complementary color relationship to the first color space. It is characterized by.

  According to a sixth aspect of the present invention, in the image processing apparatus according to the fourth or fifth aspect, the evaluation value calculating means is a weighting that is set in advance for each color component in the color intensity of the one color component. Means for calculating the saturation evaluation value based on a value multiplied by the value.

  A seventh aspect of the invention relates to a digital camera, and is characterized by including the image processing device according to any one of the first to sixth aspects.

  The image processing apparatus according to the present invention obtains a saturation evaluation value of a pixel signal in a peripheral region including a target pixel in a previous stage of interpolation processing, and suppresses a signal value of a color difference signal when the peripheral region is a low saturation region. .

  Since the saturation is evaluated from the pixel signal (RAW image data) before performing the pixel interpolation, it is possible to evaluate the saturation before the influence of noise is expanded by the interpolation. Thereby, it is possible to accurately evaluate the saturation and execute appropriate color suppression.

  In addition, since saturation is evaluated from pixel signals having linear characteristics before color matrix processing and gamma correction processing are performed, saturation can be evaluated with high accuracy.

  Furthermore, the saturation is evaluated using pixels in the peripheral area that is in the vicinity of the target pixel. As a result, whether or not to suppress color noise or the reference saturation evaluation for determining the level of suppression of color noise is calculated based on the information about the region near the target pixel that is the processing target pixel. It is possible to accurately remove pixel color noise.

<1. Overview of overall configuration and processing of digital camera>
Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a digital camera 1 according to the embodiment.

  The digital camera 1 includes a color filter array 11, a CCD (Charge Coupled Devices) 12, an A / D converter 13, an interpolation circuit 14, a matrix circuit 15, a gamma correction circuit 16, a color space conversion circuit 17, a contour enhancement / noise suppression circuit. 18, a memory 19, a saturation evaluation circuit 20, and the like.

  The CCD 12 including the color filter array 11 having the RGB Bayer array enters a subject image through an optical system (not shown) to form a subject image. The analog pixel signal output from the CCD 12 is converted into a digital pixel signal by the A / D converter 13. In addition to the CCD, a CMOS sensor can be used as the image sensor. Further, as the color filter array, a complementary color filter or the like may be used.

  Since the color filter array 11 is an RGB Bayer array, the pixel signal output from the A / D converter 13 is a signal having one of RGB color components per pixel. This pixel signal is interpolated by the interpolation circuit 14 and converted into a pixel signal having all the RGB color components per pixel.

  The pixel signal output from the interpolation circuit 14 is subjected to color correction processing for improving color reproducibility in the matrix circuit 15. The pixel signal output from the matrix circuit 15 is subjected to gamma correction processing in the gamma correction circuit 16. The pixel signal output from the gamma correction circuit 16 undergoes color space conversion from the RGB color space to the YCbCr color space in the color space conversion circuit 17.

  The pixel signal converted into the YCbCr color space by the color space conversion circuit 17 is then output to the contour enhancement / noise suppression circuit 18. The contour enhancement / noise suppression circuit 18 performs processing for enhancing the contour of the captured image and enhancing the resolution of the captured image. The contour enhancement / noise suppression circuit 18 suppresses color noise included in the pixel signal. The contour enhancement / noise suppression circuit 18 particularly improves the quality of the captured image by suppressing color noise in the gray region.

  The contour enhancement / noise suppression circuit 18 includes a contour extraction circuit 101 and an adder 102. The contour extraction circuit 101 receives the luminance signal Y and extracts a contour component from the luminance signal Y.

  The contour component signal output from the contour extraction circuit 101 is input to the adder 102. The adder 102 adds the contour component signal to the luminance signal Y and outputs a luminance signal Y in which the contour component is emphasized.

  The contour emphasis / noise suppression circuit 18 also includes low pass filters (LPF) 103 and 104. The low-pass filter 103 receives the color difference signal Cb output from the color space conversion circuit 17 and removes high frequency noise from the color difference signal Cb. The low pass filter 104 receives the color difference signal Cr output from the color space conversion circuit 17 and removes high frequency noise from the color difference signal Cr.

  Due to a decrease in light receiving sensitivity due to an increase in the total number of pixels of the CCD 12, the pixel signal includes color noise. In particular, the color noise in the gray area has a great influence on the quality of the captured image, and thus needs to be removed. The color noise in the gray area is removed by a subsequent color noise suppression processing unit. Therefore, the low-pass filters 103 and 104 may set the cut frequency to such an extent that the sharpness of the image is not lowered.

  The edge emphasis / noise suppression circuit 18 also includes look-up tables (LUTs) 105 and 106 and multipliers 107 and 108. The look-up tables 105 and 106 and the multipliers 107 and 108 are a characteristic part of the present invention together with the saturation evaluation circuit 20 and constitute a color noise suppression processing unit. The saturation evaluation circuit 20 performs arithmetic processing using the pixel signal of the pixel of interest and its surrounding area, and performs the saturation evaluation of the area around the pixel of interest. Details of the color noise suppression processing will be described later.

  The color difference signals Cb and Cr output from the low pass filters 103 and 104 are output to the memory 19 after being subjected to color noise suppression processing.

  The image signal composed of the luminance signal Y and the color difference signals Cb and Cr stored in the memory 19 is displayed on, for example, a monitor (not shown) provided in the digital camera 1. Alternatively, the image signal composed of the luminance signal Y and the color difference signals Cb and Cr is subjected to encoding processing and then stored in a memory card (not shown) attached to the digital camera 1.

<2. Notation of pixels in Bayer array>
Next, the description method of the pixels in the Bayer array in the following description and drawings will be described. First, pixels in a 5 × 5 matrix area are represented as shown in FIG. The symbol P in FIG. 2A is a notation that does not consider which color component the pixel is RGB. In contrast, in FIG. 2B to FIG. 2E, the color components of each pixel are distinguished from each other. Symbol R indicates a red pixel, symbol G indicates a green pixel, and symbol B indicates a blue pixel. 2B to 2E, the G pixel is drawn with a solid circle, and the R pixel and the B pixel are drawn with a broken circle.

  Of the subscripts of symbols P, R, G, and B, the first digit indicates the row number of the pixel in the matrix region, and the second digit indicates the column number of the pixel in the matrix region. FIGS. 2A to 2E show pixel arrangements in a matrix area composed of 25 pixels P00 to P44 including the target pixel P22. The notation method in other drawings is also the same. In the description of the embodiment and each mathematical expression, the symbols P, R, G, and B may represent pixel values. For example, the symbol P11 represents the pixel itself of the first row and the first column and also represents the pixel value of the pixel of the first row and the first column.

  FIG. 2B and FIG. 2E are pixel arrays when the target pixel P22 is a G pixel. FIG. 2C shows a pixel arrangement when the target pixel P22 is an R pixel. FIG. 2D shows a pixel arrangement when the target pixel P22 is a B pixel. As described above, the saturation evaluation circuit 20 accumulates pixel signals in the matrix region in the register group in order to perform arithmetic processing using the pixel signals of the pixel of interest and the surrounding regions. When pixels in a 5 × 5 matrix region are to be processed, there are four patterns of pixel signals stored in the register group, as shown in FIGS. 2B to 2E. Further, when processing a pixel in a 3 × 3 matrix region, nine pixels P11, P12, P13, P21, P22, P23, P31, P32, and P33 centered on the target pixel P22 are used. Accordingly, the pattern of the pixel signal is similarly the four patterns of FIGS. 2B to 2E.

<3. Saturation evaluation processing>
Next, the configuration of the saturation evaluation circuit 20 will be described. FIG. 3 is a circuit block diagram of the saturation evaluation circuit 20.

  The saturation evaluation circuit 20 includes a pixel storage unit 210. The pixel storage unit 210 stores the RGB pixel signals of the Bayer array output from the A / D converter 13. That is, the pixel storage unit 210 accumulates RAW data before pixel interpolation. The pixel storage unit 210 includes a pixel register 210a including nine register groups that store a 3 × 3 pixel matrix. In addition, the pixel storage unit 210 includes a line memory for storing pixel signals.

  The saturation evaluation circuit 20 includes average value calculation circuits 211, 212, and 213 and color intensity calculation circuits 221, 222, and 223. The saturation evaluation circuit 20 includes a comparison circuit 231, an evaluation value calculation circuit 232, and a set value storage register 233.

  The set value storage register 233 stores weight values Rw, Gw, Bw, Cw, Mw, and Yw used for calculation in the saturation evaluation circuit 20. The weighting value Rw is a weighting value for the red component. The weighting value Gw is a weighting value for the green color component. The weighting value Bw is a weighting value for the blue component. The weighting value Cw is a weighting value for the cyan color component. The weighting value Mw is a weighting value for the magenta color component. The weighting value Yw is a weighting value for the yellow component.

  In the present embodiment, as shown in Equation (1), weight values Rw, Gw, Bw, Cw, Mw, and Yw are set.

  These weighting values are coefficients that are multiplied when the saturation is evaluated. By adjusting this weighting value, the saturation can be evaluated with different characteristics for each color component. As shown in the equation (1), in this embodiment, the weight value Yw is set small. As described above, the yellow component is adjusted so that the weighting value is small and the saturation evaluation value CV is small.

  The average value calculation circuit 211 stores the green pixel G among the nine pixels stored in the pixel register 100a. The average value calculation circuit 211 calculates a pixel average value of the green pixel G.

  The average value calculation circuit 212 stores the red pixel R among the nine pixels stored in the pixel register 100a. The average value calculation circuit 212 calculates the pixel average value of the red pixel R.

  The average value calculation circuit 213 stores the blue pixel B among the nine pixels stored in the pixel register 100a. The average value calculation circuit 213 calculates the pixel average value of the blue pixel B.

  As described above, the nine matrix region pixel patterns stored in the pixel register 100a are the four patterns shown in FIGS. 2B to 2E. However, FIG. 2B to FIG. 2E illustrate 25 pixels around the pixel of interest, but the pixel register 100a stores nine pixels centered on the pixel of interest P22. Pixels (P11, P12, P13, P21, P22, P23, P31, P32, P33).

  When the pixel pattern stored in the pixel register 100a is the pattern of FIG. 2B, the average value calculation circuit 211 calculates the average pixel value G_s of the G component by the calculation shown in Equation (2). Is done. In the average value calculation circuit 212, the average pixel value R_s of the R component is calculated by the calculation shown in the equation (3). In the average value calculation circuit 213, the average pixel value B_s of the B component is calculated by the calculation shown in Equation (4).

  When the pixel pattern stored in the pixel register 100a is the pattern shown in FIG. 2C, the average value calculation circuit 211 calculates the average pixel value G_s of the G component by the calculation shown in Equation (5). Is done. In the average value calculation circuit 212, the average pixel value R_s of the R component is calculated by the calculation shown in the equation (6). In the average value calculation circuit 213, the average pixel value B_s of the B component is calculated by the calculation shown in Equation (7).

  When the pixel pattern stored in the pixel register 100a is the pattern shown in FIG. 2D, the average value calculation circuit 211 calculates the average pixel value G_s of the G component by the calculation shown in Equation (8). Is done. Further, in the average value calculation circuit 212, the average pixel value R_s of the R component is calculated by the calculation shown in Equation (9). In the average value calculation circuit 213, the average pixel value B_s of the B component is calculated by the calculation shown in Equation (10).

  When the pixel pattern stored in the pixel register 100a is the pattern shown in FIG. 2E, the average value calculation circuit 211 calculates the average pixel value G_s of the G component by the calculation shown in Equation (11). Is done. Further, in the average value calculation circuit 212, the average pixel value R_s of the R component is calculated by the calculation shown in the equation (12). In the average value calculation circuit 213, the average pixel value B_s of the B component is calculated by the calculation shown in Equation (13).

  The average value calculation circuit 211 outputs the calculated average pixel value G_s to the color intensity calculation circuits 221, 222, and 223. Similarly, the average value calculation circuits 212 and 213 also output the calculated average pixel values R_s and B_s to the color intensity calculation circuits 221, 222 and 223.

  The color intensity calculation circuit 221 uses the average value calculation circuits G_s, R_s, and B_s to calculate the color intensity difference value Gd as shown in Equation (14). Similarly, the color intensity calculation circuits 222 and 223 use the average value calculation circuits G_s, R_s, and B_s to calculate the color intensity difference values Rd and Bd as shown in Equation (14).

  Further, the color intensity calculation circuit 221 calculates a color intensity evaluation value Gda, which is an absolute value of the calculated color intensity difference value Gd, as shown in Equation (15). Similarly, the color intensity calculation circuits 222 and 223 calculate color intensity evaluation values Rda and Bda, which are absolute values of the calculated color intensity difference values Rd and Bd, as shown in Equation (15). The function abs (z) in the number (15) color indicates a function for calculating the absolute value of z.

  The color intensity calculation circuit 221 outputs the calculated color intensity difference value Gd and the color intensity evaluation value Gda to the comparison circuit 231. Further, the color intensity calculation circuit 221 outputs the color intensity evaluation value Gda to the evaluation value calculation circuit 232. Similarly, the color intensity calculation circuits 222 and 223 output the calculated color intensity difference values Rd and Bd and color intensity evaluation values Rda and Bda to the comparison circuit 231. The color intensity calculation circuits 222 and 223 output the color intensity evaluation values Rda and Bda to the evaluation value calculation circuit 232.

  The comparison circuit 231 first determines whether or not the color intensity evaluation values Gda, Rda, and Bda satisfy the relationship shown in the equation (16). When the relationship shown in the equation (16) is satisfied, the color intensity evaluation value Rda is larger than the other color intensity evaluation values Gda and Bda. It can be judged.

  When the comparison circuit 231 satisfies the relationship of the equation (16) and determines that the color intensity difference value Rd <0, the comparison circuit 231 determines that the intensity of the red R is the highest, and the evaluation value calculation circuit 232 An instruction signal S indicating that R is the highest intensity is output.

  When the instruction value S indicating that the red color R has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in Equation (17). Here, the function Div (x, y) indicates a function that divides x by y and outputs an integer part thereof. Therefore, the equation (17) means that the value obtained by multiplying the color intensity evaluation value Rda by the weight value Rw of red R is divided by 16, and the integer part is output as the saturation evaluation value CV. Division by the numerical value 16 means normalization. As shown in the equation (1), a numerical value of 16 or less is set as each weighting value. By normalizing by division of the numerical value 16, the saturation evaluation value CV can be evaluated low for colors having a weighting value smaller than 16. In the calculations of the following formulas (18), (20), (21), (23), and (24), division by the numerical value 16 indicates normalization.

  When the comparison circuit 231 satisfies the relationship of the equation (16) and determines that the color intensity difference value Rd ≧ 0, the comparison circuit 231 determines that the intensity of the cyan color C is the highest, and the evaluation value calculation circuit 232 An instruction signal S indicating that the cyan color C has the highest intensity is output.

  When the instruction value S indicating that the cyan color C has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in Equation (18). In equation (18), the value obtained by multiplying the color intensity evaluation value Rda by the weighting value Cw of the cyan color C is divided by the numerical value 16, and the integer part is output as the saturation evaluation value CV.

  If the comparison circuit 231 determines that the relationship expressed by the equation (16) is not satisfied, it then determines whether the relationship indicated by the equation (19) is satisfied. When the relationship shown in the equation (19) is satisfied, the color intensity evaluation value Gda is larger than the other color intensity evaluation values Rda and Bda. It can be judged.

  When the comparison circuit 231 satisfies the relationship of the equation (19) and determines that the color intensity difference value Gd <0, the comparison circuit 231 determines that the intensity of green G is the highest, and causes the evaluation value calculation circuit 232 to display the green color. An instruction signal S indicating that G is the highest intensity is output.

  When the instruction value S indicating that the green G has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in the equation (20). That is, the value obtained by multiplying the color intensity evaluation value Gda by the weight value Gw of green G is divided by the numerical value 16, and the integer part is output as the saturation evaluation value CV.

  When the comparison circuit 231 satisfies the relationship of the equation (19) and determines that the color intensity difference value Gd ≧ 0, the comparison circuit 231 determines that the intensity of the magenta color M is the highest, and the evaluation value calculation circuit 232 An instruction signal S indicating that the magenta color M has the highest intensity is output.

  When the instruction value S indicating that the magenta color M has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in Equation (21). That is, the value obtained by multiplying the color intensity evaluation value Gda by the weighting value Mw of the magenta color M is divided by the numerical value 16, and the integer part is output as the saturation evaluation value CV.

  If the comparison circuit 231 determines that neither of the expressions (16) and (19) is satisfied, then the comparison circuit 231 determines whether or not the relationship represented by the expression (22) is satisfied. When satisfying the relationship shown in the equation (22), the color intensity evaluation value Bda is larger than the other color intensity evaluation values Gda and Rda, so that the intensity of yellow Y, which is blue B or a complementary color of blue B, is biased. I can judge.

  When the comparison circuit 231 satisfies the relationship of the equation (22) and determines that the color intensity difference value Bd <0, the comparison circuit 231 determines that the intensity of blue B is the highest, and the evaluation value calculation circuit 232 An instruction signal S indicating that B is the highest intensity is output.

  When the instruction value S indicating that blue B has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in Equation (23). That is, the value obtained by multiplying the color intensity evaluation value Bda by the weight value Bw of blue B is divided by the numerical value 16, and the integer part is output as the saturation evaluation value CV.

  If the comparison circuit 231 satisfies the relationship of the equation (22) and determines that the color intensity difference value Bd ≧ 0, the comparison circuit 231 determines that the intensity of yellow Y is the highest, and the evaluation value calculation circuit 232 An instruction signal S indicating that Y is the highest intensity is output.

  When the instruction value S indicating that yellow Y has the highest intensity is input, the evaluation value calculation circuit 232 calculates the saturation evaluation value CV by performing the calculation shown in Equation (24). That is, the value obtained by multiplying the color intensity evaluation value Bda by the weight value Yw of yellow Y is divided by the numerical value 16, and the integer part is output as the saturation evaluation value CV. As described above, the weighting value Yw is set to a smaller value than other color components. Thereby, in this Embodiment, it adjusts so that the color noise of a yellow component may be removed easily. This is because the color noise of the yellow component existing in the gray region is recognized particularly conspicuously.

  If the comparison circuit 231 determines that none of the relations of the expression (16), the expression (19), and the expression (22) is satisfied, then the comparison circuit 231 determines whether the relation of the expression (25) is satisfied. judge.

  The case where the relationship of the equation (25) is satisfied is a case where Rd = Gd = Bd = 0, that is, the peripheral region of the target pixel is a complete gray region. When this condition is satisfied, the comparison circuit 231 outputs an instruction signal S indicating that the peripheral area is a complete gray area to the evaluation value calculation circuit 232. The evaluation value calculation circuit 232 receives the instruction signal S, and sets the saturation evaluation value CV to 0 as shown in the equation (26).

  When the comparison circuit 231 determines that none of the relations of Expression (16), Expression (19), Expression (22), and Expression (25) is satisfied, the relation of Expression (27) is then given. It is determined whether or not the above is satisfied.

  If any one of the three expressions shown in the expression (27) is satisfied, the comparison circuit 231 further determines whether or not Rda is 0. Then, the comparison circuit 231 outputs an instruction signal S indicating that there are a plurality of color components with the highest intensity and indicating whether Rda is 0 to the evaluation value calculation circuit 232.

  The evaluation value calculation circuit 232 sets the saturation evaluation value CV by the equation (28) if Rda is 0 when the instruction signal S indicating that there are a plurality of color components having the highest intensity is input.

  When the instruction signal S indicating that the highest intensity color component is plural is input to the evaluation value calculation circuit 232, if Rda is not 0, the saturation evaluation value CV is set by the equation (29).

  When satisfying any one of the three formulas shown in the equation (27), the two color intensity evaluation values take the highest intensity value, and the remaining one color intensity evaluation value is zero. Therefore, in the case of Rda = 0, Gda or Bda takes the same maximum intensity value, and therefore the value of one Gda is set as the saturation evaluation value CV. Of course, the value of Bda may be set to the saturation evaluation value CV. In addition, since the two color intensity evaluation values have the highest intensity, the color is not largely biased to a single color component, so that the weighting value is not multiplied.

  On the other hand, when Rda = 0 is not true, Rda takes the maximum intensity value together with any other color intensity evaluation value. Therefore, Rda is set as the saturation evaluation value CV. Of course, the same is true even if the color intensity evaluation value of the other highest intensity color component is set. Also in this case, since the bias to a specific color is not large, weighting values are not multiplied.

  The evaluation value calculation circuit 232 is an arithmetic expression of any one of formulas (17), (18), (20), (21), (23), (24), (26), (28), and (29). When the saturation evaluation value CV is calculated by the above, finally, the calculation shown by the equation (30) is performed, and the saturation evaluation value CV is further normalized.

  In the equation (30), fm is a normalization parameter. The normalization parameter fm is stored in the set value storage register 233. As the normalization parameter fm, for example, numerical values such as 1, 2, 4, 8, 16, 32, 64, and 128 can be selected and set. The evaluation value calculation circuit 232 outputs the saturation evaluation value CV normalized by the normalization parameter fm to the contour enhancement / color noise suppression circuit 18.

<4. Color noise suppression processing>
Please refer to FIG. 1 again. The saturation evaluation value CV calculated by the saturation evaluation circuit 20 is output to the look-up tables 105 and 106. Look-up tables 105 and 106 calculate suppression coefficients Kb and Kr from the saturation evaluation value CV. The suppression coefficient Kb is a coefficient that suppresses the signal of the color difference signal Cb, and the suppression coefficient Kr is a coefficient that suppresses the signal of the color difference signal Cr.

  FIG. 4 is a diagram illustrating the relationship between the saturation evaluation value CV and the suppression coefficient Kb that are associated in the lookup table 105. As shown in the figure, when the saturation evaluation value CV is greater than or equal to a certain threshold value q, 1 is associated with the suppression coefficient Kb. That is, when the saturation evaluation value CV is large, the color difference signal Cb is not suppressed. On the other hand, when the saturation evaluation value CV is smaller than the threshold value q and is determined to be a low saturation region, a value smaller than 1 is associated with the suppression coefficient Kb. For example, the suppression coefficient Kb = α (<1) is associated with the saturation evaluation value CV = p (<q).

  Although the relationship shown in FIG. 4 has been described as the correspondence relationship in the lookup table 105, the relationship between the saturation evaluation value CV and the suppression coefficient Kr associated in the lookup table 106 is the same. However, by setting different correspondences between the saturation evaluation value CV and the suppression coefficient in the lookup tables 105 and 106, respectively, it is possible to have different characteristics for each color difference component.

  Look-up tables 105 and 106 output suppression coefficients Kb and Kr to multipliers 107 and 108, respectively. The multiplier 107 suppresses the color difference signal Cb by multiplying the color difference signal Cb input from the low-pass filter 103 by the suppression coefficient Kb. The multiplier 108 suppresses the color difference signal Cr by multiplying the color difference signal Cr input from the low-pass filter 104 by the suppression coefficient Kr.

  FIG. 5 is a diagram showing the suppression level of the color difference signal Cb. In the figure, the horizontal axis represents the signal value of the color difference signal Cb input by the multiplier 107, and the vertical axis represents the signal value of the color difference signal Cb after color suppression output from the multiplier 107. When the saturation evaluation value CV is equal to or greater than the threshold value q, as described above, since the suppression coefficient Kb = 1, the input signal value and the output signal value of the multiplier 107 are the same. In other words, the color is not suppressed in the region where the saturation evaluation is high. Thereby, the vividness of the region with high saturation can be maintained.

  On the other hand, when the saturation evaluation value CV is smaller than the threshold value q, the output signal value is suppressed. For example, when the suppression coefficient α (CV = p in FIG. 4), the color difference signal Cb of the signal value d input by the multiplier 107 is output as a signal value α · d (<d).

  Although the relationship shown in FIG. 5 has been described as the relationship between the signal value of the color difference signal Cb input by the multiplier 107 and the signal value of the color difference signal Cb output by the multiplier 107, the relationship between the color difference signal Cr input by the multiplier 108 is described. The relationship between the signal value and the signal value of the color difference signal Cr output from the multiplier 108 is the same. However, the suppression level can be made different for each color difference component depending on the characteristics of the lookup tables 105 and 106.

  As described above, in the digital camera 1 according to the present embodiment, the saturation around the target pixel is evaluated, and the color components of the color difference signals Cb and Cr are suppressed in the low saturation area. The generated color noise can be suppressed.

  In particular, in the present embodiment, since the saturation is evaluated from the pixel signal (RAW image data) before pixel interpolation, it is possible to evaluate the saturation before expanding the influence of noise by interpolation. . Thereby, it is possible to accurately evaluate the saturation and execute appropriate color suppression.

  In addition, since saturation is evaluated from pixel signals having linear characteristics before color matrix processing and gamma correction processing are performed, saturation can be evaluated with high accuracy.

  Then, the digital camera 1 according to the present embodiment evaluates the saturation using pixels in the peripheral region that is the vicinity of the target pixel (in this embodiment, nine pixels around the target pixel). As a result, whether or not to suppress color noise or the reference saturation evaluation for determining the level of suppression of color noise is calculated based on the information about the region near the target pixel that is the processing target pixel. It is possible to accurately remove pixel color noise.

  In the present embodiment, the normalization parameter (the numerical value 16 in this embodiment) used in the equations (17), (18), (20), (21), (23), and (24) is used. By adjusting, the scale after normalization of the saturation evaluation value CV can be adjusted for each color component. Furthermore, the scale of the final saturation evaluation value CV can be adjusted by adjusting the normalization parameter fm used in Equation (26).

1 is a block diagram of a digital camera according to an embodiment. It is a figure which shows the color pattern in the matrix area | region of the pixel signal of a Bayer arrangement. It is a block diagram of a saturation evaluation circuit. It is a figure which shows a response | compatibility with the saturation evaluation value and suppression coefficient which the lookup table hold | maintains. It is a figure which shows the color suppression level of a color difference signal.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Digital camera 18 Contour emphasis / noise suppression circuit 20 Saturation evaluation circuit 105,106 Lookup table 107,108 Multiplier CV Saturation evaluation value Gd, Rd, Bd Color intensity difference value Gda, Rda, Bda Color intensity evaluation value

Claims (7)

Input means for inputting a pixel signal of a first color space having one color component per pixel;
A saturation evaluation means for obtaining a saturation evaluation value of a pixel signal in a peripheral region including the target pixel input by the input means;
Interpolation means for interpolating the pixel signal input by the input means to generate a pixel signal having all color components of the predetermined color space for one pixel;
Color space conversion means for converting the pixel signal of the first color space interpolated by the interpolation means into a pixel signal of a second color space including a luminance signal and a color difference signal;
By comparing the saturation evaluation value with a predetermined condition, it is determined whether the surrounding area is a low saturation area, and if it is determined that the surrounding area is a low saturation area, the color difference Suppression means for suppressing the signal value of the signal;
An image processing apparatus comprising: The image processing apparatus according to claim 1.
The suppression means includes
Multiplying means for multiplying the color difference signal by a suppression coefficient;
Coefficient determination means for holding a correspondence relationship between the saturation evaluation value and the suppression coefficient, and determining the suppression coefficient from the saturation evaluation value;
An image processing apparatus comprising: The image processing apparatus according to claim 2,
The color difference signal is
A first color difference signal and a second color difference signal;
Including
The coefficient determining means includes
First coefficient determining means for maintaining a correspondence relationship between the saturation evaluation value and the first suppression coefficient corresponding to the first color difference signal, and determining the first suppression coefficient from the saturation evaluation value;
Second coefficient determining means for maintaining a correspondence relationship between the saturation evaluation value and a second suppression coefficient corresponding to the second color difference signal, and determining the second suppression coefficient from the saturation evaluation value;
Including
The multiplication means is
First multiplication means for multiplying the first color difference signal by the first suppression coefficient;
Second multiplying means for multiplying the second color difference signal by the second suppression coefficient;
An image processing apparatus comprising: The image processing apparatus according to any one of claims 1 to 3,
The saturation evaluation means includes
Color intensity calculating means for calculating the color intensity of each color component in the first color space;
A color component determining means for comparing each color intensity and determining one color component having the highest color intensity based on the degree of bias of each color intensity;
Evaluation value calculating means for calculating the saturation evaluation value based on the color intensity of the one color component;
An image processing apparatus comprising: The image processing apparatus according to claim 4.
The color component determining means includes
The color intensity of each color component of the first color space calculated by the color intensity calculation means is compared, and the color component included in the first color space or the complementary color relationship with the first color space is obtained. Means for determining the one color component from among the color components included in a certain color space;
An image processing apparatus comprising: The image processing apparatus according to claim 4 or 5,
The evaluation value calculation means includes
Means for calculating the saturation evaluation value based on a value obtained by multiplying the color intensity of the one color component by a weighting value set in advance for each color component;
An image processing apparatus comprising:   A digital camera comprising the image processing apparatus according to claim 1.

Images