JP2001078226A - Method and device for image generation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To output a suitable image corresponding to its purpose, photography scene, etc., by selecting >=3 channels corresponding to a target image from a multichannel image of >=4 channels, generating different kinds of 3-channel images, and putting them together into a 3-channel image. SOLUTION: An image acquisition part 12 acquires a multichannel image of >=4 channels from an image data supply source R such as a digital camera. An image generation part 10 obtains a 4-channel image composed of four source images of Y(yellow), G(green), C(cyan), and M(magenta) and generates a 3- channel image of R, G, and B from the 4-channel image. An image generation part 14 generates a 3-channel image of R, G, and B by a user's choice or in response to a signal for image generation instruction such as picture quality adjustment instruction by using source images for maintaining the three channels.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention belongs to the technical field of image creation for creating a three-channel image from image data of a multi-channel image.

[0002]

2. Description of the Related Art Generally, photographing of a color image using an image sensor such as a CCD sensor used in a digital camera, a scanner, or the like is performed using red (R), blue (G), and green (B).
, And three subtractive primary colors of C (cyan), M (magenta), and Y (yellow).

On the other hand, in recent years, a multi-channel image pickup system having four or more channels has been put into practical use for the purpose of improving color reproduction accuracy and the like. For example, in a multi-channel imaging system called a color difference sequential system, a single plate (one CCD sensor) is formed by using a color filter array as shown in FIG.
However, a four-channel imaging system of Y, M, C and G is realized.

In such a color difference sequential method, when creating R, G, and B three-channel images corresponding to output from the obtained four-channel image data,
Generally, this is performed using a matrix of 3 rows and 4 columns (3 × 4).

For example, Japanese Patent Application Laid-Open No. 10-98735 discloses a device having a color filter array (CCD) shown in FIG.
Sensor), the device signals (Y, G, C, M) of four pixels adjacent to each other are used to calculate 3 × 4
Luminance signal (C _L) and a color difference signal (C _r, C _b) a determined using the _{matrix, C L = Y + G +} C + M C r = from Y-G-C + M C b = -Y-G + C + M The luminance and color difference signals, the following formula R, G and B signals are obtained using a 3 × 3 matrix shown by _{R = (C L + 4C r} -C b) / 10 G = (2C L -2C r -2C b) / 10 B = (C L -C r + 4C b) / 10

[0006]

However, in the conventional multi-channel imaging system including the above-described method,
The effect of multi-channeling is only the improvement of color reproduction accuracy, and it cannot be said that a sufficient effect is obtained with respect to an increase in cost due to multi-channeling.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art, and not only to improve the color reproduction accuracy, but also to exhibit the advantages unique to a multi-channel image pickup system. Intent, purpose of image and shooting scene (picture)
It is an object of the present invention to provide an image creating method and an image creating apparatus capable of outputting a suitable image according to the above.

[0008]

In order to achieve the above object, an image forming method according to the present invention selects a plurality of channels by selecting three or more channels corresponding to a target image from a multi-channel image of four or more channels. Create a 3 channel image of
There is provided an image creating method characterized by creating a three-channel image by combining a plurality of created three-channel images.

In the image forming method according to the present invention, it is preferable that the combination of the plurality of three-channel images is based on a linear combination, or that a weighting coefficient of the linear combination is set according to a target image. .

[0010] The image creating apparatus according to the present invention further comprises: an image acquiring unit for acquiring a multi-channel image of four or more channels;
An image creation unit for creating a plurality of types of three-channel images from the multi-channel images acquired by the image acquisition unit using three or more channels according to an image creation instruction, and a plurality of three-channel images created by the image creation unit And an image synthesizing unit for synthesizing the image.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an image forming method and an image forming apparatus according to the present invention will be described in detail based on preferred embodiments shown in the accompanying drawings.

FIG. 1 is a block diagram showing an example of an image forming apparatus of the present invention for implementing the image forming method of the present invention.
The image creation device 10 in the illustrated example basically includes the image acquisition unit 1
2, an image creation unit 14, an image synthesis unit 16, and a weight calculation unit 18.

The image acquiring section 12 acquires a multi-channel image (image data) of four or more channels from an image data source R such as a digital camera. As an example, the image creating apparatus 10 shown in FIG.
A four-channel image composed of four original images of G (green), C (cyan) and M (magenta) is obtained, and three channels of R (red), G and B (blue) are obtained from the four-channel image. (3ch) Create an image (image data).

In the present invention, the multi-channel image obtained from the image data supply source R may be a multi-channel image obtained by a single-plate type image pickup means such as the color difference sequential method described above. Alternatively, it may be a multi-channel image obtained by a multi-plate type imaging means in which light carrying an image is divided into four parts by using a prism, a half mirror, or the like, and captured by four CCD sensors. .

The image creating section 14 converts the multi-channel image acquired by the image acquiring section 12 into three or more channels in response to a signal S of an image creating instruction such as a user selection or an image quality adjusting instruction (weighting in each mode). Using the original image, R, G and B three-channel images are created.

As described above, Y,
G, C and M four channel images are provided. When R, G, and B three-channel images (hereinafter, simply referred to as three-channel images) are created from such four-channel images using original images of three or more channels, five types of three-channel images are used in combination. Although channel images can be created, according to the study of the present inventors, these three channel images have different characteristics depending on the combination of the original images.

More specifically, when a three-channel image is created by using all the four-channel original images of Y, G, C, and M, the color reproduction is excellent in balance and, in particular, 1
In the case of a single-panel imaging system in which one channel is assigned to a pixel, an image with good sharpness in image structure can be obtained. When a three-channel image is created using three-channel original images of Y, C, and M, the color reproduction is excellent in balance, and the image structure is reduced in S / N ratio due to the G original image. And a low-noise image with good graininess can be obtained. Also, as is clear from the hue circle,
When a three-channel image is created using the three-channel original images of G, C, and M, an image with good color reproduction of the cool color system can be obtained, and the three-channel original images of Y, G, and M are used. Conversely, when a three-channel image is created, an image with good warm color reproduction can be obtained.
When a three-channel image is created using three-channel original images of Y, G, and C, an image with good green color reproduction can be obtained.

That is, when an image having excellent color balance and sharpness is intended, it is preferable to create an image using all four-channel original images. In the case of an image in which noise is a rate-limiting factor for image quality, G is not used, and Y, C, and M are not used.
It is preferable to create an image using only the three channel original images. Further, for example, in the case of an image in which most of the image is occupied by flesh color, such as a close-up photograph of a person, a warm color system is used by using three-channel original images of Y, G, and M excluding C, which is a complementary color of flesh color. It is preferable to create an image with good color reproduction, and for landscape photos with many green areas,
It is preferable to create an image with good green color reproduction by using the original image of three channels of Y, G and C without using M.

According to the present invention, the characteristics of a multi-channel image (multi-channel image pickup system) can be sufficiently achieved to customize the photographing function according to the purpose, to perform specialized photographing according to the use of the image, It is possible to output a suitable image according to the photographing purpose and photographing intention, the use of the image, the photographing scene (pattern), and the like.

A plurality of modes, for example, various modes such as a high-sharpness mode, a low-noise mode, a cool color mode, a warm color mode, and a green mode are set in the image forming apparatus 10 in the illustrated example. A plurality of combinations are selected according to the use of the subject or the image. The selection of these modes is instructed to the image creating unit 14 by the input signal S of the image creating instruction.

The image creating section 14 has the following modes according to each mode.
A 3 × 4 matrix of M _{0 to} M ₄ shown below for creating a three-channel image using three or more channels from the four-channel image supplied to the image acquisition unit 12 is set, and the designated mode is set. And processes the four-channel image.

[0022]

(Equation 1)

The matrix M ₀ is for creating a three-channel image by using all the four-channel original images of Y, G, C and M. The image creating section 14 selects the high sharpness mode. in this case, the image acquisition unit 12 to create a 3-channel image by processing the 4-channel image obtained by using the matrix M _0. The matrix M ₁ is
Y terms are all 0, that is,
G, is intended to create a 3-channel image using only original image C and M, the image creation unit 14, when the cold color mode is selected, the image acquisition unit using the matrix M ₁ 12 processes the acquired 4-channel image to create a 3-channel image. Hereinafter, similarly, when the low noise mode is selected, the image creating unit 14
G terms are all 0 Y, the matrix M ₂ using only the original image of C and M, when the warm mode is selected, all C of the term that is a 0 Y, G, and M of the matrix M ₃ using only the original image, further, if the greenish mode is selected, the section M are all 0 Y, the matrix M ₄ using only original image G and C, respectively using The four-channel image acquired by the image acquisition unit 12 is processed to create a three-channel image.

In the present invention, a plurality of such three-channel images are created from one multi-channel image, and a three-channel image obtained by combining these is output. In this case, the combination is preferably performed by linear combination, and more preferably, the weighting coefficient k of the linear combination is adjusted according to a target image. By performing such a combination, it is possible to further improve the degree of freedom in creating an image and to output a more suitable image according to the purpose of capturing the image, the purpose of capturing the image, the image, and the like.

In the illustrated image forming apparatus 10, the image forming unit 14 forms a plurality of three-channel images using a plurality of matrices in accordance with an input signal S.
The weight calculator 18 calculates the weighting coefficients of the three-channel images created by the respective matrices, and
In step 6, the plurality of created three-channel images are linearly combined according to the weighting coefficients and combined to form a three-channel image to be output.

For example, in response to an instruction from the user, a mode for adjusting the balance between sharpness and noise (or selecting both the high sharpness mode and the low noise mode) is selected, and the image quality is adjusted (for example, the user In the case of performing an adjustment instruction with a higher degree, the image creating unit 14 converts the four-channel image supplied to the image acquiring unit 12 using the matrix M ₀ and the matrix M ₂ according to the input signal S. By performing the processing, two three-channel images are created, and the weight calculator 18 calculates the weight coefficient k (0 <k <
1) is set, and the image synthesizing unit 16 linearly combines the two three-channel images using the weight coefficient k to create a three-channel image for output. M (k) = k · M ₀ + (1−k) · M _{2 In} other words, by defining a new matrix using the weight coefficient k (0 <k <1), the balance between sharpness and noise can be improved. Can be changed.

Similarly, for example, using a matrix M ₀ realizing high sharpness, a matrix M ₁ excellent in cold color reproducibility, and a matrix M ₃ excellent in warm color reproducibility, respective weighting factors k ₀ , k ₁ and k ₃ (k ₀ + k ₁ + k ₃ =
By defining a new matrix using 1), a mode in which color reproduction can be customized can be set. M (k ₀ , k ₁ , k ₃ ) = k ₀ · M ₀ + k ₁ · M ₁ + k
₃ · M ₃

In the present invention, the weighted synthesis of a plurality of three-channel images is performed as in the above-described example.
The weight of the image in each matrix may be set by the user, or the mode may be a mode in which the weight of the image in each matrix is set in advance, and both may be selectable depending on the mode.

In the present invention, the image forming unit 14
Is not limited to forming a three-channel image from a multi-channel image by selecting a matrix to be used. For example, an image created by using all matrices is basically synthesized, and unnecessary By setting the weight coefficient k to 0 for an image, an original image used for creating a three-channel image may be selected.

Although the image forming method and the image forming apparatus of the present invention have been described in detail above, the present invention is not limited to the above-described example, and various changes and improvements can be made without departing from the gist of the present invention. Of course, you can do it.

[0031]

Hereinafter, the present invention will be described in more detail with reference to specific examples of the present invention.

380 nm, 430 nm, 480 nm, 5
30 nm, 580 nm, 630 nm, 680 nm, 73
Using the image data of a multiband image captured by inserting nine optical filters having peaks at 0 nm and 780 nm, respectively, and the spectral sensitivity characteristics shown in FIG. 2, a Y original image, a G original image, Four-channel image (image data) consisting of four original images of C original image and M original image
It was created. The four-channel image is divided into a color difference sequential method (single-plate method) using the filter arrangement shown in FIG. 3 and a four-sheet method in which four channels are assigned to each pixel.
Created ones for each. In addition, a noise whose amplitude is inversely proportional to the parallel root of the sensitivity (integrated value of spectral sensitivity) of each channel is superimposed on each of the four channel images.

The Y, G, C, and M four-channel images created in this manner are combined with the matrices M _{0 to} M ₄ described above.
To produce three-channel images (R, G and B). As a result, even in any of the 4-channel image of a single plate type and 4-plate system, three-channel image color balance and sharpness created in the matrix M ₀ is 3 channel image created by the matrix M ₁ is cold color color reproducibility of the three-channel image created by the matrix M ₂ the color balance and graininess, created in the matrix M ₃ 3
Channel image color reproduction of warm colors, the three-channel image created by the matrix M ₄ color reproducibility of green system, it was confirmed that each feature point.

In the present invention, a plurality of three-channel images are created by selecting a matrix corresponding to a desired image, and are synthesized using the matrix. For example, when strobe light is prohibited as in an art museum, or when strobe light is set weaker to avoid unnaturalness due to strobe light, the obtained image is due to lack of sensitivity. Noise (granularity) is noticeable. In this case, the image created by the matrix M ₀ and the matrix M ₂
It was found that a good image could be obtained by combining the images created in the above. Further, the shooting environment, from the balance of sharpness and noise, optimal image is sometimes located between the image created by the image and the matrix M ₂ created in the matrix M _0. In this case, from the photometric value and the shutter speed and aperture, and the image created by the matrix M _0, it sets the internal ratio of the images created by the matrix M _2, to synthesize the two images according to the interior division ratio Output was found to be good. In addition, the landscape, such as most or limited to skin color, if you want to faithfully reproduce the skin color, the image by the image and the matrix M ₃ by the matrix M _0, mountains and forests in the subject as in the case of shooting up of a person Is taken, an image based on the matrix M ₀ and the matrix M
₄ an image by, in the case of shooting the scene of the coast, the image by the image and the matrix M ₁ by the matrix M _0, it has been found that may be respectively synthesized. further,
Etc. The group photo multiplayer, the image by the matrix M _0, and the image by the matrix M _2, it was found that an image of the matrix M ₃ may be synthesized. From the above results, the effect of the present invention is clear.

[0035]

As described in detail above, according to the present invention, not only the color reproduction accuracy can be improved, but also the advantages unique to a multi-channel imaging system can be exerted well, and the customization and use of the photographing function can be achieved. According to specialization, etc., it is possible to output a suitable image according to the purpose and purpose of capturing the image, the purpose of the image, the shooting scene, and the like.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of an image creating apparatus according to the present invention using an image creating method according to the present invention.

FIG. 2 is a graph showing spectral characteristics in an example of the present invention.

FIG. 3 is a diagram illustrating an example of a filter array in photographing a multi-channel image of a color difference sequential system.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Image creation apparatus 12 Image acquisition part 14 Image creation part 16 Image synthesis part 18 Weight calculation part

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 9/64 F term (Reference) 5B057 BA02 BA12 BA25 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CC01 CE08 CE17 5C057 AA01 AA06 DA04 EA00 EA01 GL02 GM01 5C065 AA01 AA03 AA07 BB01 BB16 BB22 CC01 DD17 DD19 DD26 EE07 5C066 AA01 AA03 CA05 CA08 EE04 GA00 GA01 5C076 AA12 AA19 AA26 BA03

Claims (3)

[Claims] 1. A multi-channel image having four or more channels, three or more channels corresponding to a target image are selected to generate a plurality of types of three-channel images, and the generated plural three-channel images are synthesized. An image creating method characterized by creating a three-channel image. 2. The method according to claim 1, wherein the synthesis of the plurality of three-channel images is based on a linear combination, or a weighting coefficient of the linear combination is set according to a target image.
Image creation method described in. 3. An image acquisition unit for acquiring a multi-channel image of four or more channels, and a plurality of three-channel images from the multi-channel images acquired by the image acquisition unit using three or more channels in accordance with an image creation instruction. An image creating apparatus comprising: an image creating unit that creates a seed; and an image combining unit that combines a plurality of three-channel images created by the image creating unit.