JP2007195258A - Color imaging apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a color imaging apparatus and a color image correcting method capable of acquiring an image of excellent image quality. <P>SOLUTION: In image data of each color (data indicating tone of each color) of each of pixels constituting an image, a value to be acquired in formula 1 becomes almost zero by randomly selecting a coefficient. As a result, a combination of image data in which a value to be acquired by the formula 1 is not almost zero is acquired for each pixel, it is considered that the image data contain noise, and the noise component is removed by correcting each image data so that the absolute value of a value indicated by the formula 1 becomes as small as possible. The formula 1 is represented as k1×d1+k2×d2+ ... +kn×dn, where n is the number of color components constituting an image, k1-kn are correcting coefficients obtained previously experimentally, and d1-dn are image data of each of colors at each of pixel positions. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a color imaging apparatus and a color image correction method capable of obtaining a high-quality image by removing noise components from image data acquired by an imaging element or the like.

  A conventional color imaging apparatus removes noise components in order to improve image quality. As a removal method, for example, a method of extracting a so-called false color portion appearing in a spatial high frequency portion or an edge portion of an image and reducing the saturation of the false color portion has been adopted.

However, according to this method, since the color of the false color portion approaches gray uniformly, there is a drawback that the color of the image after removing the noise component deviates from the actual color. In addition, in order to prevent a malfunction in which a portion that is not a false color portion is regarded as a false color and lightens the color, it is necessary to provide a dead zone, and the noise component may not be completely removed.
Furthermore, according to this method, a noise component that appears due to electrical noise or the like cannot be removed.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide a color imaging apparatus and a color image correction method capable of appropriately removing noise components and obtaining a high-quality image.

In order to achieve the above object, a color imaging apparatus according to a first aspect of the present invention includes input means for inputting image data for n colors at each pixel position or adjacent pixel position;
Correction means for correcting the value of the image data at each pixel position so that the image data input by the input means makes the absolute value of the value represented by Formula 5 as small as possible;
Means for outputting the image data corrected by the correcting means.

[Equation 5]
k1 · d1 + k2 · d2 + ... + kn · dn
n: number of color components constituting the image k1 to kn: correction coefficient d1 to dn: image data of each color at each pixel position

  Note that the correction does not require the absolute value of Formula 5 to be the minimum value, and the correction may be terminated when it is determined that the absolute value is almost the minimum value according to a predetermined criterion or the like.

  The image data d <b> 1 to dn may be corrected so that the value represented by Expression 5 becomes approximately zero. In this case as well, the correction may be terminated when it is determined that the value has become substantially zero according to a predetermined standard or the like.

By decomposing the color image into n color components and assuming that the image data at each pixel position or the image data at adjacent pixel positions is d1 to dn, the correction coefficient k1 to kn is appropriately selected to obtain Equation 5 The value obtained in is almost 0 regardless of the image. If the value of Equation 5 does not become almost zero, it is considered that the image data includes noise, and the absolute value of the value of Equation 5 is made as small as possible, or the value of Equation 5 becomes almost zero. As described above, the noise included in the image data can be removed by correcting the image data.
Therefore, according to the present invention, it is possible to remove a noise component included in image data before correction and obtain a high-quality image.

  The input unit includes an imaging unit that captures an image of a subject to generate the image data, and the correction unit is obtained based on image data obtained by irradiating the imaging unit with light of a plurality of colors in advance. Coefficient information storage means for storing the correction coefficients k1 to kn may be provided.

  The correction unit may correct the value of each image data so that the absolute value of the value represented by Expression 6 is as small as possible for the image data input by the input unit.

[Equation 6]
(D1′−d1) ² + (d2′−d2) ² +... + (Dn′−dn) ² + w (k1 · d1 ′ + k2 · d2 ′ +... + Kn · dn ′) ²
n: Number of color components constituting the image k1 to kn: Correction coefficient w: Positive weight constant d1 to dn: Image data of each color at each pixel position before correction d1 ′ to dn ′: At each pixel position after correction Image data for each color

  Note that the correction does not require the absolute value of the expression 6 to be the minimum value, and the correction may be terminated when it is determined that the absolute value is almost the minimum value in accordance with a predetermined standard or the like.

  The image data d1 to dn may be corrected so that the value represented by Expression 6 is approximately zero. In this case as well, the correction may be terminated when it is determined that the value has become substantially zero in accordance with a predetermined standard or the like.

  According to the present invention, Formula 6 is used instead of Formula 5 for correcting the image data. When Expression 6 is used, it is possible to suppress the correction amount of each image data as compared with the case where Expression 5 is used, and further, it is possible to perform correction processing in consideration of correction weights.

  The input means inputs image data formed by arranging image data for n colors in a matrix, and the correction means causes the image data input by the input means to make Equation 6 as small as possible. In addition, the value of the image data at each pixel position may be corrected.

  The correction means further includes coefficient information storage means for storing the correction coefficients k1 to kn obtained based on image data obtained by irradiating the imaging means with a plurality of colors of light in advance and the weight positive constant w. May be.

  The imaging means may further include a plurality of light receiving elements arranged in a matrix and n color filters arranged in a matrix on the light receiving surface of each light receiving element.

  According to the present invention, the image pickup means includes a plurality of light receiving elements arranged in a matrix and n color filters arranged in a matrix on the light receiving surface of each light receiving element. Input image data.

  The color filter is composed of, for example, four color filter pieces of white, green, light blue, and yellow.

  According to the present invention, there is provided a color filter comprising a plurality of light receiving elements arranged in a matrix and filter pieces of, for example, white, green, light blue, and yellow arranged in a matrix on the light receiving surface of each light receiving element. Image data of each color of each pixel is input by the provided imaging means.

  The input means may further comprise interpolation means for obtaining image data for n colors at each pixel position from image data obtained by each light receiving element.

  According to the present invention, n-color image data at the same pixel position can be input from each pixel obtained by the light receiving element by the interpolation means.

The color image correction method according to the second aspect of the present invention is:
Input image data of each color at each pixel position,
Correct the value of the input image data so that the absolute value of the value represented by Formula 5 or 6 is as small as possible,
The corrected image data is output.

  According to the present invention, the noise component is removed by correcting the input image data of each color at each pixel so that the value represented by Equation 5 or Equation 6 is approximately 0, thereby improving the image quality. Images can be acquired.

  As described above, according to the present invention, a high-quality image can be obtained.

A color camera imaging apparatus and a color image correction method according to embodiments of the present invention will be described below with reference to the drawings.
First, correction of image data performed in this embodiment will be described.
When image data of each color of each pixel constituting the image (or image data of each color of a plurality of adjacent pixels) is d1 to dn, the value obtained by Equation 7 can be obtained by appropriately selecting the correction coefficients k1 to kn. Is almost zero. This relationship holds regardless of the spectral intensity distribution of the incident light.

[Equation 7]
k1 · d1 + k2 · d2 + ... + kn · dn
n: number of color components constituting the image k1 to kn: correction coefficient d1 to dn: image data of each color at the same pixel position

Accordingly, for each pixel, when a set of image data in which the value of Equation 7 is not substantially zero is obtained, the image data is considered to contain noise, and each image data is represented by the absolute value represented by Equation 7. The noise component can be removed by correcting so that the value is as small as possible or the value represented by Equation 7 is approximately zero.
Therefore, in this embodiment, correction coefficients k1 to kn are experimentally obtained in advance using light of various spectral intensity distributions, and the image data of the actually captured image is reduced as much as Equation 7 becomes possible. By correcting, the noise component included in the image data is removed.

  Next, an embodiment of an electronic camera device using the above correction will be described with reference to FIGS.

  As shown in FIG. 1, the electronic camera device includes a shutter button 1, a control circuit 2, an imaging lens 3, a color filter 4, an imaging element 5, a memory 6, and an image display circuit 7. The

The shutter button 1 is composed of a push button or the like, and generates a trigger signal instructing the start of a series of imaging operations by a push operation.
The imaging lens 3 is made of a transparent material such as glass or plastic, and light that has passed through the imaging lens 3 forms an image on the light receiving surface of the imaging element 5.

  The color filter 4 is arranged in front of the image sensor 5 (on the imaging surface) and is composed of a color filter piece made of cellulose, gelatin or the like. The color filter piece has a pass band characteristic centered on any one of white (Wh), green (G), light blue (Cy), and yellow (Ye), and is in a matrix form as shown in FIG. Is arranged. The adjacent color filter pieces are shielded from light by a black mask.

  The imaging element 5 is composed of a CCD or the like, and has a configuration in which light receiving elements (photoelectric conversion elements) having openings as shown in FIG. 2 are arranged in a matrix, in response to an instruction from the control circuit 2. A signal corresponding to the amount of light incident on each light receiving element is output as digital image data.

  The memory 6 is composed of a RAM or the like, and stores image data output from the image sensor 5, image data processed by the control circuit 2, and determinant coefficients described later.

  The control circuit 2 is composed of an MPU (microprocessor unit) or the like, and (1) controls the imaging by the imaging device 5 in response to the pressing operation of the shutter button 1, and (2) the image data obtained by the imaging. Γ correction is performed so as to be proportional to the amount of light. (3) Interpolation processing is performed to convert the image data after γ correction into image data corresponding to each color of the color filter 4 at each pixel position. (4) Performs correction processing to remove noise components from the interpolated image data. (5) R (red), G (green), B (blue) for outputting (displaying, storing, printing) the corrected image data. Processing for converting into luminance data of each color is performed, and (6) display processing for displaying the processed image data on the display circuit 7 is executed.

Next, correction of image data performed by the electronic camera device of this embodiment will be described.
As shown in FIG. 2, the image sensor 5 of the electronic camera device of this embodiment has a configuration in which the color filter 4 is white (Wh), green (G), light blue (Cy), and yellow (Ye). Therefore, Expression 7 is expressed by Expression 8.

[Equation 8]
a ・ Wh + b ・ G + c ・ Cy + d ・ Ye
Wh to Ye: Image data output from the light receiving unit of the image sensor 5 upon receiving the light after passing through the white, green, light blue, and yellow filter pieces a, b, c, d: correction coefficients

  The light that has passed through the color filter 4 becomes light in a band centered on the color of each filter piece (white (Wh), green (G), light blue (Cy), and yellow (Ye)). When light in a band centered on the color of each filter piece is incident on the image sensor 5 and the image data output from the image sensor 5 is expressed as Wh to Ye, correction coefficients a, b, c, and d are set appropriately. The value obtained by Equation 8 is approximately 0 by selecting.

  The correction coefficients a, b, c, and d are selected experimentally because they change depending on the combination of the color filter 4 and the image sensor 5. In order to select correction coefficients a, b, c, and d experimentally, for example, the image sensor 5 is irradiated with light of a plurality of colors having different frequency bands as shown in FIG. Measure. For example, when the correction coefficients a, b, c, d are selected as a = 1, b = 0.935, c = −1.044, d = −0.915 from the measured image data Wh to Ye, As shown in FIG. 5, the value of Expression 8 is almost zero.

Formula 8 can also be used for the relationship between the image data Wh to Ye for four colors at the same pixel position. However, as shown in FIG. 2, since the color filter 4 in which four color filter pieces are arranged in a matrix is used, at each pixel position of the image sensor 5, as shown in FIG. Only the minute image data Wh to Ye can be obtained. For this reason, image data Wh to Ye that cannot be obtained at each pixel position are obtained using interpolation processing. FIG. 4 shows the obtained image data as Wh ′ _{i, j to} Ye ′ _{i, j} .

By correcting the image data Wh ′ _{i, j to} Ye ′ _{i, j} obtained in this way so that the value represented by Equation 8 is as small as possible, in other words, approximately 0, Noise components and the like can be removed.

  The correction does not necessarily require that the absolute value of the value of Formula 8 be the minimum value or 0, but when it is determined that the absolute value is almost the minimum value or the value is almost 0 according to a predetermined standard or the like. The correction may be terminated.

In addition, Equation 8 may be used to correct the image data. However, if Equation 8 is used as it is, the correction may be excessive and the balance of each color at each pixel position may be disturbed. Is used, which can be a minimum value.
Also in Equation 9, the corrected image data Wh ′ from the image data Wh ′ _{i, j to} Ye ′ _{i, j} before correction so that the absolute value of the evaluation number P _{i, j is} as small as possible or almost zero. “ _{I, j to} Ye” _{i, j} is obtained. By bringing the evaluation number P _{i, j} closer to 0, the deviation between the uncorrected image data Wh ′ _{i, j to} Ye ′ _{i, j} and the corrected image data Wh ″ _{i, j to} Ye ″ _{i, j} Can be small.

[Equation 9]
P _{i, j} = (Wh ″ _{i, j} −Wh ′ _{i, j} ) ² + (G ″ _{i, j} −G ′ _{i, j} ) ² + (Cy ″ _{i, j} −Cy ′ _{i, j} ) ² + (Ye ″ _{i, j} −Ye ′ _{i, j} ) ² + w · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c · Cy ″ _{i, j} + d · Ye ″ _{i, j} ) ²
Wh ′ _{i, j to} Ye ′ _{i, j} : White, green, light blue, yellow image data before correction Wh ″ _{i, j to} Ye ″ _{i, j} : White, green, light blue, yellow image data after correction P _{i, j} : Evaluation number w: Weight positive constant a, b, c, d: Correction coefficient

In order to reduce the absolute value of the evaluation number P _{i, j in} Equation 9 as much as possible, the relationship between Wh ′ _{i, j to} Ye ′ _{i, j} and Wh ″ _{i, j to} Ye ″ _{i, j} is corrected. The image data Wh ″ _{i, j to} Ye ″ _{i, j} can be expressed by a determinant based on the image data Wh ′ _{i, j to} Ye ′ _{i, j} before correction. By using this determinant, the corrected image data Wh ″ _{i, j to} Ye ″ _{i, j} can be directly obtained from the uncorrected image data Wh ′ _{i, j to} Ye ′ _{i, j} . An example of the procedure is shown below.
First, as shown in Formula 10, Formula 9 is partially differentiated by Wh ″ _{i, j to} Ye ″ _{i, j} .

[Equation 10]
∂P _{i, j} / ∂Wh ″ _{i, j} = 2 · (Wh ″ _{i, j} −Wh ′ _{i, j} ) + 2 · w · a · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c・ Cy ″ _{i, j} + d ・ Ye ″ _{i, j} )
∂P _{i, j} / ∂G ″ _{i, j} = 2 · (G ″ _{i, j} −G ′ _{i, j} ) + 2 · w · b · (aWh ″ _{i, j} + b · G ″ _{i, j} + c · Cy “ _{I, j} + d · Ye” _{i, j} )
∂P _{i, j} / ∂Cy ″ _{i, j} = 2 · (Cy ″ _{i, j} −Cy ′ _{i, j} ) + 2 · w · c · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c・ Cy ″ _{i, j} + d ・ Ye ″ _{i, j} )
∂P _{i, j} / ∂Y ″ _{i, j} = 2 · (Ye ″ _{i, j} −Ye ′ _{i, j} ) + 2 · w · d · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c・ Cy ″ _{i, j} + d ・ Ye ″ _{i, j} )

Next, as shown in Equation 11, Equation 10 is a quaternary simultaneous equation with Wh ″ _{i, j to} Ye ″ _{i, j} as variables.

[Equation 11]
2 · (Wh ″ _{i, j} −Wh ′ _{i, j} ) + 2 · w · a · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c · Cy ″ _{i, j} + d · Ye ″ _{i, j} ) = 0
2 · (G ″ _{i, j} −G ′ _{i, j} ) + 2 · w · b · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c · Cy ″ _{i, j} + d · Ye ″ _{i, j} ) = 0
2 · (Cy ″ _{i, j} −Cy ′ _{i, j} ) + 2 · w · c · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c · Cy ″ _{i, j} + d · Ye ″ _{i, j} ) = 0
2 · (Ye ″ _{i, j} −Ye ′ _{i, j} ) + 2 · w · d · (a · Wh ″ _{i, j} + b · G ″ _{i, j} + c · Cy ″ _{i, j} + d · Ye ″ _{i, j} ) = 0

  Next, Formula 11 is organized into a determinant as shown in Formula 12.

  Equation 12 sets the correction coefficients a, b, c, and d to, for example, a = 1, b = 0.935, c = −1.044, d = −0.915 obtained from FIG. For example, when the constant is selected as w = 5.0, it can be expressed by Expression 13.

Ｗｈ’_i,j〜Ｙｅ’_i,j：補正前画像データ
Ｗｈ”_i,j〜Ｙｅ”_i,j：補正後画像データ

Wh ′ _{i, j to} Ye ′ _{i, j} : Image data before correction Wh ″ _{i, j to} Ye ″ _{i, j} : Image data after correction

That is, based on the image data Wh ′ _{i, j to} Ye ′ _{i, j} , the corrected image data Wh ″ _{i, j to} Ye ″ _{i, j} can be directly obtained by Expression 13.
Note that the coefficients of the determinant of Expression 13 are stored in the memory 6 in advance.

Next, the imaging operation of the electronic camera device will be described with reference to the flowchart of FIG.
When the shutter button 1 is pressed, the shutter button 1 gives a trigger signal for instructing the control circuit 2 to start imaging.
In response to the trigger signal, the control circuit 2 controls the image sensor 5 to perform exposure and output image data.
Since the image data output from the image sensor 5 includes a smear component and a dark current value, these are removed as shown in Equation 14.

[Formula 14]
D ′ _Wh = D _Wh −D _D −D _S
D ′ _G = D _G −D _D −D _S
D ′ _Cy = D _Cy −D _D −D _S
D ' _Ye = D _Ye -D _D -D _S
D _{Wh to} D _Ye : Image data of white, green, light blue and yellow before removal D ′ _{Wh to} D ′ _Ye : Image data of white, green, light blue and yellow after removal D _D : Component due to dark current D _S : Smear ingredient

Since the light receiving element of the image pickup element 5 has the γ characteristic as shown in FIG. 7A, the value of the output image data is not proportional to the amount of light incident on the pixels of the light receiving element. Therefore, as shown in Expression 15, the γ characteristic indicated by the image sensor 5 is obtained in advance as a function, and image data is substituted into the inverse function γ ⁻¹ (x) to perform γ correction for canceling the γ characteristic (step) S1).
[Equation 15]
D ' _Wh = γ _Wh ^-1 (D _Wh )
D ′ _G = γ _G ⁻¹ (D _G )
D ' _Cy = γ _Cy ^-1 (D _Cy )
D ' _Ye = γ _Ye ^-1 (D _Ye )
D _{Wh to} D _Ye : White, green, light blue, yellow image data before γ correction D ′ _{Wh to} D ′ _Ye : White, green, light blue, yellow image data after γ correction γ ⁻¹ : Indicates a light receiving element Inverse function of γ characteristic function

  As shown in FIG. 7B, the value of the image data output from the light receiving element is proportional to the amount of light incident on the pixel of the light receiving element, as shown in FIG.

At each pixel position of the image pickup device 5, as shown in FIGS. 2 and 3, only image data Wh to Ye for one color exists. Therefore, the control circuit 2 uses Wh ′ to Ye ′, which are image data for four colors at each pixel position, from the image data after γ correction, as shown in FIG. Obtained by interpolation processing (step S2).
The interpolation processing also serves as a (spatial) low-pass filter, and can improve the image quality of the image data Wh ′ to Ye ′ for four colors at each pixel position.

[Equation 16]
Wh ' _{i, j} = (Wh _{i-2, j-2} + Wh _{i-2, j} + Wh _{i-2, j + 2} + Wh _{i, j-2} + Wh _{i, j} + Wh _{i, j + 2} + Wh _{i + 2, j-2} + Wh _{i + 2, j} + Wh _{i + 2, j + 2} ) / 9
(When Wh _{i, j} exists)
G ′ _{i, j} = (G _{i−2, j−2} + G _{i−2, j} + G _{i−2, j + 2} + G _{i, j−2} + G _{i, j} + G _{i, j + 2} + G _{i + 2, j-2} + G _{i + 2, j} + G _{i + 2, j + 2} ) / 9
(When G _{i, j} exists)
Cy ' _{i, j} = (Cy _{i-2, j-2} + Cy _{i-2, j} + Cy _{i-2, j + 2} + Cy _{i, j-2} + Cy _{i, j} + Cy _{i, j + 2} + Cy _{i + 2, j-2} + Cy _{i + 2, j} + Cy _{i + 2, j + 2} ) / 9
(When Cy _{i, j} exists)
Ye ' _{i, j} = (Ye _{i-2, j-2} + Ye _{i-2, j} + Ye _{i-2, j + 2} + Ye _{i, j-2} + Ye _{i, j} + Ye _{i, j + 2} + Ye _{i + 2, j-2} + Ye _{i + 2, j} + Ye _{i + 2, j + 2} ) / 9
(If Ye _{i, j} exists)
Wh _{i, j to} Ye _{i, j} : White, green, light blue, yellow pre-interpolation image data Wh ′ _{i, j to} Ye ′ _{i, j} : White, green, light blue, yellow post-interpolation image data

[Equation 17]
Wh ′ _{i, j} = (Wh _{i−1, j−1} + Wh _{i−1, j + 1} + Wh _{i + 1, j + 1} + Wh _{i + 1, j + 1} ) / 4
(When Wh _{i, j} does not exist)
G ′ _{i, j} = (G _{i−1, j−1} + G _{i−1, j + 1} + G _{i + 1, j + 1} + G _{i + 1, j + 1} ) / 4
(When G _{i, j} does not exist)
Cy ′ _{i, j} = (Cy _{i−1, j−1} + Cy _{i−1, j + 1} + Cy _{i + 1, j + 1} + Cy _{i + 1, j + 1} ) / 4
(When Cy _{i, j} does not exist)
Ye ′ _{i, j} = (Ye _{i−1, j−1} + Ye _{i−1, j + 1} + Ye _{i + 1, j + 1} + Ye _{i + 1, j + 1} ) / 4
(If Ye _{i, j} does not exist)
Wh _{i, j to} Ye _{i, j} : White, green, light blue, yellow pre-interpolation image data Wh ′ _{i, j to} Ye ′ _{i, j} : White, green, light blue, yellow post-interpolation image data

For example, the control circuit 2, when obtaining the image data Wh _{'2, 2} shown in FIG. 4, since the image data Wh _{2, 2} exists as shown in FIG. 3, fitting the equation 16, Wh _{2, 2} And eight adjacent image data Wh _4,0 , Wh _2,0 , Wh _0,0 , Wh _4,2 , Wh _0,2 , Wh _4,4 , Wh _2,4 , Wh _0,4 Image data Wh ' ₂ , ₂ is obtained. When obtaining the image data Wh ′ _1,1 shown in FIG. 4, since the image data Wh _1,1 does not exist as shown in FIG. 3, Equation 17 is applied, and four adjacent to Ye _1,1 are obtained. Average image data Wh ′ _1,1 is obtained from the image data Wh _2,0 , Wh _0,0 , Wh _2,2 , Wh _0,2 . Similarly, corrected interpolation data for all pixel positions is obtained.
The number of obtained image data is four times the number of original image data. The control circuit 2 stores the obtained image data in the memory 6.

  After the interpolation processing, the control circuit 2 performs correction processing according to Equation 13 using the image data whose number has been quadrupled and a determinant coefficient stored in advance in the memory 6 to obtain corrected image data. (Step S3).

For example, the control circuit 2 determines the determinant previously obtained experimentally stored in the memory 6 based on the interpolated image data Wh ′ _{0,0 to} Ye ′ _0,0 stored in the memory 6. Is used to perform correction processing to obtain corrected image data Wh ″ _{0,0 to} Ye ″ _0,0 . Similarly, the control circuit 2 obtains corrected image data for all pixel positions.
By the correction process, the image data is corrected based on Expression 13, and the relationship of Expression 8 is substantially satisfied. Accordingly, it is possible to acquire a high-quality image from which noise components are removed from the image data. The control circuit 2 stores the image data from which the noise component has been removed in the memory 6.

  After the correction processing, the control circuit 2 uses R (red), G (green), B (blue) using, for example, Equation 18 so that the image data from which the noise component has been removed can be used for output (display, printing). ) Processing for converting into luminance data of each color is performed (step S4).

Ｒ〜Ｇ：Ｒ（赤）、Ｇ（緑）、Ｂ（青）の輝度
ａ１１〜ａ３４：変換係数
Ｗｈ〜Ｙｅ：白、緑、水色、黄色の画像データ

R to G: luminance of R (red), G (green), and B (blue) a11 to a34: conversion coefficient Wh to Ye: white, green, light blue, yellow image data

  The control circuit 2 displays the image data on the image display circuit 7 based on the RGB luminance data obtained by the processing (step S5).

  As described above, according to this embodiment, since the image data is corrected so as to make Equation 8 as small as possible, a high-quality image from which noise components are removed can be obtained.

  In the electronic camera device of the above embodiment, interpolation processing using Formula 16 and Formula 17 is performed, but in that case, peripheral pixels used in Formula 16 and Formula 17 may be increased, and a separate coefficient is used. May be added.

  In the above embodiment, the image data Wh ′ to Ye ′ for a plurality of colors at each pixel position is acquired. However, the image data of each color component at each pixel position may be acquired using other methods.

For example, an optical path switch as shown in FIG. 8 may be disposed in front of the color filter 4 (on the imaging surface) to acquire all the image data Wh to Ye for four colors at each pixel position.
The optical path switch includes a polarizing plate 8, a first liquid crystal plate 9, a first refracting plate 10, a second liquid crystal plate 11, and a second refracting plate 12.

The polarizing plate 8 changes incident light into linearly polarized light.
The first liquid crystal plate 9 and the second liquid crystal plate 11 are formed of a pair of transparent substrates having stripe-shaped electrodes formed on opposite surfaces and subjected to alignment treatment in a direction shifted by 90 °, and 90 ° according to the alignment treatment. TN (Twisted Nematic) liquid crystal sealed between transparent substrates with a twist angle of The direction of the alignment process of the transparent substrate on the light incident side of the first liquid crystal plate 9 is set parallel to the transmission axis of the polarizing plate 8 and the direction of the alignment process of the transparent substrate on the light incident side of the second liquid crystal plate 11. Is set perpendicular to the transmission axis of the polarizing plate 8.
The first refracting plate 10 and the second refracting plate 12 are made of, for example, a birefringent crystal such as a quartz plate or calcite, and are normal light (light in a vibration direction perpendicular to the plane including the crystal optical axis) and abnormal. It has a different refractive index depending on the light. The first refracting plate 10 and the second refracting plate 12 adjust the refractive index and the thickness. When the incident light is ordinary light, the first refracting plate 10 goes straight. When the incident light is abnormal light, the first refracting plate 10 and the second refracting plate 12 are equivalent to one pixel pitch. Set to shift. The optical axis of the first refracting plate 10 is set so that its horizontal component is perpendicular to the transmission axis of the polarizing plate 8, and the optical axis of the second refracting plate 12 has its horizontal component as polarizing plate 8. Is set in a direction parallel to the transmission axis.

Hereinafter, acquisition of image data Wh to Ye for four colors at each pixel position using the optical path switch will be described with reference to FIGS. 9 and 10.
The control circuit 2 controls the voltages applied to the first liquid crystal plate 9 and the second liquid crystal plate 11 to (1) straightly advance the incident light, and (2) one pixel pitch of the image sensor 5. 2 (3) 1 pixel pitch of the image sensor 5 is shifted in the column direction of FIG. 2, and (4) 1 pixel pitch of the image sensor 5 is shifted in the row direction and the column direction of FIG. Shift.

For example, as shown in FIG. 10, light that is originally incident only on the position Wh _0,0 of the color filter 4 is first applied with a voltage to the first liquid crystal plate 9 and the second liquid crystal plate 11 at the first timing. Is not applied (FIG. 9A-1), and image data is obtained at the position Wh _0,0 of the color filter 4.

Next, at a second timing, a voltage is not applied to the first liquid crystal plate 9, and a voltage is applied to the second liquid crystal plate 11 (FIG. 9 (a-2)). Image data is obtained by shifting the optical path to the position G _0,1 (row direction) of the color filter 4.

Next, at a third timing, a voltage is applied to the first liquid crystal plate 9 and the second liquid crystal plate 11 (FIG. 9B-1), and the position Cy _{1,0 of the} color filter 4 is set. Image data is obtained by shifting the optical path in the (column direction).

Finally, at the fourth timing, the voltage is applied to the first liquid crystal plate 9 and the voltage is not applied to the second liquid crystal plate 11 (FIG. 9 (b-2)). Image data is obtained by shifting the optical path to the position Ye _1,1 (row direction and column direction) of the color filter 4. The image data obtained in this way can be image data of Wh _0,0 , G _0,0 , Cy _0,0 , Ye _0,0 , respectively.
By doing the same at other pixel positions, it is possible to acquire all the image data Wh to Ye for four colors at each pixel position.

  The optical path switch controls the voltage applied to the first liquid crystal plate 9 and the second liquid crystal plate 11 and shifts the incident light. However, the color filter 4 and the image sensor 5 are moved relative to each other to shift the incident light. You may let them.

  In the above embodiments, the electronic camera device has been described. However, the imaging device may be a video camera device, a facsimile device, a scanner device, or a copy device.

  Further, the corrected image data may be directly calculated from Expression 8 instead of using Expression 9.

  Further, the color of the color filter (color component constituting the image) is not limited to Wh (white), G (green), Cy (light blue), and Ye (yellow), and can be arbitrarily changed. For example, a combination of Y (yellow), M (magenta), C (cyan), R (red), G (green), and B (blue) may be used as the color filter.

  Further, using a rotary color filter or the like, a Y (yellow) color filter is arranged on the front surface of the image sensor 5 to obtain Y (yellow) image data, and an M (magenta) color filter is arranged. Thus, M (magenta) image data is obtained, and a C (cyan) color filter is arranged to obtain C (cyan) image data. is there.

  In addition, the imaging device for acquiring image data and the control circuit (correction circuit) for correcting the obtained image data do not need to be configured integrally, and are imaged at other locations and stored in the medium. The image data may be corrected by a correction circuit separate from the imaging device to remove noise.

In addition, the example in which the image data of each color at the same pixel position is corrected using Equation 8 or Equation 9 has been mainly described. However, the image data of adjacent pixels of different colors makes Equation 8 or Equation 9 as small as possible. You may correct to. For example, for each combination of Wh _{2i, 2j} , G _{2i, 2j + 1} , Cy _{2i + 1,2j} , Ye _{2i + 1,2j + 1} (i and j are 0 or natural numbers) of the image data as shown in FIG. Correction processing may be performed.
Furthermore, after correcting image data of adjacent pixels, adjustment may be performed between a plurality of adjacent sets.

Furthermore, the method of selecting the correction coefficients k1 to kn of Expression 7 and the correction coefficients a to d of Expression 8 is arbitrary.
For example, as described above, the image sensor 5 is actually irradiated with light of a plurality of colors having different spectral intensity distributions, and the obtained image data (image after removing dark current and smear components and performing γ correction) Data may be selected).

  Moreover, when an image is actually acquired by imaging, a coefficient that satisfies most of the obtained image data in Expression 8 or Expression 9 may be selected as necessary. Further, the coefficient may be selected logically.

In the above embodiment, the same number of image data for the four color filters at the same position as the original image data is calculated. However, the number of image data is less than this, for example, 1/4 in both the vertical and horizontal directions (that is, one color). It may be calculated by using a calculation method such as a digital low-pass filter as many as the number of original pixel data for the color filter).
The present invention mainly affects the color (difference) data. Since the resolution of the human eye for color (difference) data is lower than the resolution for luminance data (monochrome data), it does not have a substantial effect.

It is a block diagram which shows an example of the specific structure of the color imaging device concerning embodiment of this invention. It is a figure which shows the arrangement | sequence of the color filter 4 concerning the embodiment of this invention, and the arrangement | sequence of the image pick-up element 5. FIG. It is a figure which shows the arrangement | sequence of the color filter 4 concerning embodiment of this invention. It is a figure which shows the arrangement | sequence of the image data Wh-Ye for four colors in each pixel position concerning embodiment of this invention. It is a figure which shows the value of the image data which the image pick-up element 5 outputs when the color filter 4 on the image pick-up element 5 concerning embodiment of this invention is irradiated with light of multiple colors. It is a flowchart which shows an example of operation | movement of the electronic camera apparatus concerning embodiment of this invention. It is a figure which shows the relationship between the light quantity and image data in the image pick-up element 5 concerning embodiment of this invention. It is a figure which shows the structural example of an optical path switch. It is a figure which shows the optical path of an optical path switch. It is a figure which shows typically the incident light irradiated to the color filter 4 on the image pick-up element 5 with an optical path switch.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Shutter button, 2 ... Control circuit, 3 ... Imaging lens, 4 ... Color filter, 5 ... Imaging element, 6 ... Memory, 7 ... Image display circuit, 8 ... Polarizing plate, 9 ... first liquid crystal plate, 10 ... first refracting plate, 11 ... second liquid crystal plate, 12 ... second refracting plate

Claims (14)

Input means for inputting image data for n colors at each pixel position;
Correction means for correcting the value of the image data at each pixel position so that the image data input by the input means makes the absolute value of the value shown in Formula 1 as small as possible;
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising:
[Equation 1]
k1 · d1 + k2 · d2 + ... + kn · dn
n: number of color components constituting the image k1 to kn: correction coefficient d1 to dn: image data of each color at each pixel position Input means for inputting image data for n colors at each pixel position;
Correction means for correcting the value of the image data at each pixel position so that the image data input by the input means satisfies Formula 2.
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising:
[Equation 2]
k1 · d1 + k2 · d2 + ... + kn · dn≈0
n: number of color components constituting the image k1 to kn: correction coefficient d1 to dn: image data of each color at each pixel position input means for inputting image data formed by arranging image data for n colors in a matrix;
Correction means for correcting the value of the image data at each adjacent pixel position so that the image data input by the input means makes the absolute value of the value represented by Equation 1 as small as possible;
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising: input means for inputting image data formed by arranging image data for n colors in a matrix;
Correction means for correcting the value of the image data at each adjacent pixel position so that the image data input by the input means satisfies Equation (2);
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising: The input means includes imaging means for capturing an image of a subject and generating the image data,
The correction means includes coefficient information storage means for storing the correction coefficients k1 to kn obtained based on image data obtained by irradiating the imaging means with a plurality of colors of light in advance.
The color imaging device according to claim 1, wherein the color imaging device is a color imaging device. Input means for inputting image data for n colors at each pixel position;
Correction means for correcting the value of the image data at each pixel position so that the image data input by the input means makes the value represented by Equation 3 as small as possible;
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising:
[Equation 3]
(D1′−d1) ² + (d2′−d2) ² +... + (Dn′−dn) ² + w (k1 · d1 ′ + k2 · d2 ′ +... + Kn · dn ′) ²
n: Number of color components constituting the image k1 to kn: Correction coefficient w: Positive weight constant d1 to dn: Image data of each color at each pixel position before correction d1 ′ to dn ′: At each pixel position after correction Image data for each color Input means for inputting image data for n colors at each pixel position;
Correction means for correcting the value of the image data at each pixel position so that the image data input by the input means satisfies Formula 4.
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising:
[Equation 4]
(D1′−d1) ² + (d2′−d2) ² +... + (Dn′−dn) ² + w (k1 · d1 ′ + k2 · d2 ′ +... + Kn · dn ′) ² ≈0
n: Number of color components constituting the image k1 to kn: Correction coefficient w: Positive weight constant d1 to dn: Image data of each color at each pixel position before correction d1 ′ to dn ′: At each pixel position after correction Image data for each color input means for inputting image data formed by arranging image data for n colors in a matrix;
Correction means for correcting the value of the image data at each adjacent pixel position so that the image data input by the input means makes the absolute value of the value represented by Equation 3 as small as possible;
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising: input means for inputting image data formed by arranging image data for n colors in a matrix;
Correction means for correcting the value of the image data at each adjacent pixel position so that the image data input by the input means satisfies Equation (4);
Means for outputting the image data corrected by the correction means;
A color imaging apparatus comprising: The input means includes imaging means for capturing an image of a subject and generating the image data,
The correction means includes coefficient information storage means for storing the correction coefficients k1 to kn obtained based on image data obtained by irradiating the imaging means with a plurality of colors of light in advance and the weight positive constant w.
The color imaging device according to claim 6, wherein the color imaging device is a color imaging device. The imaging means includes a plurality of light receiving elements arranged in a matrix, and n color filters arranged in a matrix on the light receiving surface of each light receiving element,
The input means includes interpolation means for obtaining image data for n colors at each pixel position from image data obtained by each light receiving element using an interpolation method;
The color imaging apparatus according to claim 1, further comprising:   The color imaging apparatus according to claim 11, wherein the color filters are four colors of white, green, light blue, and yellow. Input image data of each color at each pixel position or adjacent pixel position,
Correcting the value of the input image data so as to make the absolute value of the value shown in Formula 1 or 3 as small as possible,
A color image correction method comprising outputting corrected image data. Input image data of each color at each pixel position or adjacent pixel position,
Correct the value of the input image data so as to satisfy Equation 2 or 4,
A color image correction method comprising outputting corrected image data.

Images