JP2001119556A - Image processor and processing method for multi-band image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor and an image processing method by which an image of an object can colorimetrically accurately be reproduced in matching with the atmosphere when the image is photographed in the case of synthesizing the object with a background image and a synthesized image can be generated by uniquely deciding a desired light source. SOLUTION: This image processing provides the image processing unit for a multi-band image and its image processing method, where a spectral intensity distribution for a desired light source is selected from a light source spectral distribution database section storing a plurality of light source spectral intensity distributions, a desired background image is selected from a background image database section storing a plurality of multi-band image data as synthesis purpose background images, synthesizing the background image spectral refractive index data obtained from the image data with object spectral refractive index data obtained from the multi-band image of the object can obtain synthesized image spectral refractive index data, and using the synthesize image spectral refractive index data and the spectral intensity distribution can obtain data of the synthesized image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-band image of a subject composed of a plurality of images obtained by dividing a photographing wavelength band into a plurality of band bands and a desired multi-band background image. The present invention relates to an image generation technique for generating a composite image under a desired light source using image data.

[0002]

2. Description of the Related Art Today, with the advance of digital image processing, a subject image photographed by a digital camera or the like can be easily pasted and synthesized on an arbitrary background image using photo retouching software. For example, a person image photographed indoors using a light source as an incandescent light bulb can be combined with a background image photographed on a summer beach to create a person image on the beach. The synthesis of such a color image
Generally, this is performed by replacing a part of the image data of the background image of each pixel of R (red), G (green) and B (blue) with the image data of the subject image. Therefore, in the obtained composite image, the colorimetric values of the subject image are maintained as they are and are reproduced together with the background image.

[0003]

However, the colorimetric value of the photographed subject image is affected by the spectral intensity distribution of the photographing light source.
It has a color temperature of 0K (Kelvin) and the image tends to be reddish. On the other hand, a background image taken on a summer beach does not have a red tint like a subject image using an incandescent bulb as a light source because the light source uses sunlight having a color temperature of 5000 to 6000 K (Kelvin) as a light source. For this reason, the composite image reproduced with the background image while maintaining the colorimetric values of the subject image as it is, the subject is separated from the background image and emerges, making it a realistic image as if shooting a person on the beach Can not get.

For example, even if the image data of each of the R, G, and B pixels is arbitrarily changed using photo retouching software to adjust the color balance so that the subject image matches the color of the background image. However, there is a problem that an image that matches the appearance atmosphere when actually photographing a subject together with a background under a desired light source cannot be colorimetrically accurately reproduced, and a sufficiently realistic composite image cannot be obtained. there were.

Therefore, the present invention solves the above problems,
When synthesizing the subject with the background image, the subject's image that matches the atmosphere at the time of shooting the image is accurately reproduced colorimetrically,
An image processing apparatus capable of fusing a subject image with a background image without a sense of incongruity and obtaining a sufficiently realistic composite image, and at the time of creating a composite image, uniquely defining a desired light source and creating a composite image; and It is an object to provide an image processing method.

[0006]

In order to achieve the above object, the present inventor, when photographing a subject, divides a photographing wavelength band into a plurality of band bands and a plurality of bands photographed for each band band. As a result of intensive studies focusing on a multi-band image composed of images, the following invention has been achieved. That is, the present invention is a multi-band image processing apparatus for processing a multi-band image composed of a plurality of images obtained by dividing a photographing wavelength band into a plurality of band bands when photographing a subject. A light source spectral distribution database unit including a plurality of spectral intensity distributions of light sources for use, a background image database unit including a plurality of multiband background images having a known spectral intensity distribution of a light source at the time of shooting to be combined as a background image of a subject, and the light source Selecting means for selecting a desired light source spectral intensity distribution and a desired multi-band background image from the spectral distribution database section and the background image database section; and a background image spectrum obtained from the multi-band background image selected by the selecting means. Reflectance data and subject spectral reflectance data obtained from a multi-band image of the subject. An image data synthesizing unit for obtaining synthesized image spectral reflectance data, and using the synthesized image spectral reflectance data obtained by the image data synthesizing processing unit and the spectral intensity distribution selected by the selecting unit, A composite image generation unit that obtains image data of the composite image, and an image data output unit that converts the image data of the composite image obtained by the composite image generation unit into a predetermined output image signal and outputs the output signal. The present invention provides an image processing apparatus for a multi-band image.

Further, the present invention provides a method for photographing a subject,
An image processing method of a multi-band image for processing a multi-band image composed of a plurality of images obtained by dividing an imaging wavelength band into a plurality of band bands, comprising: a light source spectral distribution database including a plurality of spectral intensity distributions of an illumination light source. A desired spectral intensity distribution of the light source for illumination from the unit, and synthesize a desired multiplicity as a background image of the subject from a background image database unit including a plurality of multi-band background images with known spectral intensity distributions of the light source at the time of shooting. Select the band background image,
Background image spectral reflectance data obtained from the selected multi-band background image is combined with subject spectral reflectance data obtained from the multi-band image of the subject to obtain composite image spectral reflectance data. Image processing of a multi-band image, wherein image data of a composite image is obtained using reflectance data and the spectral intensity distribution, and the image data of the composite image is converted into a predetermined output image signal and output. It provides a method.

Here, the light source spectral distribution database section has a spectral intensity distribution corresponding to the color temperature of the outdoor light source,
When an outdoor light source is selected as a desired illumination light source, at least one of a place, a season, and a time zone is designated as an inspection key, the color temperature of the light source is determined by using the search key, and the spectral intensity of the light source is determined from the color temperature. Preferably, the distribution is determined. At that time, when selecting a desired spectral intensity distribution of the illumination light source, a reference function that defines the color temperature of the outdoor light source with respect to time is set according to the specified location. Then, the reference key is compressed / expanded at a ratio corresponding to the specified season, and the color temperature in the specified time zone is obtained from the compressed / expanded reference function to obtain the search key. More preferably, the color temperature is determined.
When obtaining the image data of the composite image, it is preferable to obtain the image data of the composite image by changing the intensity of the light source used for the background image region and the subject image region in the image data of the composite image.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a multi-band image processing apparatus of the present invention for implementing a multi-band image processing method of the present invention will be described in detail based on a preferred embodiment shown in the accompanying drawings.

FIG. 1 shows a multi-band image processing apparatus (hereinafter, referred to as a multi-band image processing apparatus) according to an embodiment of the present invention.
FIG. 1 shows a schematic block diagram of an apparatus (referred to as the present apparatus) 10. This device 1
0 may be constituted by software in which the functions described below are processed by a computer, or some or all of the functions may be constituted by an independent device. The present apparatus 10 pastes a subject image of a multi-band image obtained by photographing a subject O illuminated by the light source S with the multi-band camera 12 onto a desired background image, synthesizes the image, and converts the image data of the synthesized image into a printer. 30 is an image processing apparatus that outputs the image data to the image processing apparatus 30.

The apparatus 10 includes a multi-band image recording unit 14 for holding a multi-band image captured by a multi-band camera 12 and a component of a spectral intensity distribution of a light source S from a multi-band image. A spectral reflectance estimating unit 16 for estimating a spectral reflectance distribution and obtaining subject spectral reflectance data, a light source spectral distribution database unit 18 having a plurality of light source spectral intensity distributions, and a multi-band background to be synthesized as a background image of the subject. A background image database unit 20 having a plurality of images and a light source spectral distribution database unit 18
A search unit 22 for selecting and searching a desired spectral intensity distribution and a multi-band background image from the background image database unit 20 and a background image spectral reflection obtained from the multi-band background image selected by the search unit 22 The image data synthesizing unit 24 that synthesizes the ratio data and the subject spectral reflectance data to obtain synthesized image spectral reflectance data, and uses the synthesized image spectral reflectance data and the spectral intensity distribution selected by the search unit 22. A composite image generation unit 26 for obtaining image data of the composite image, and an image data output unit 28 for converting the image data of the composite image obtained by the composite image generation unit 26 into a predetermined output image signal and outputting the signal. It is configured to have. In addition, the apparatus 10 is connected to the printer 3 via the image data output unit 28 in order to print out the composite image.
Connected to 0.

The multi-band camera 12 is a camera for photographing a subject O to photograph a multi-band image composed of a plurality of band images obtained by dividing a photographing wavelength band into a plurality of band bands. Filter as a band-pass filter that divides light into a plurality of band bands, and C that captures transmitted light transmitted through the variable filter
It consists of a CD camera. The obtained image data is sent to the multiband image recording unit 14 of the present apparatus 10 and is temporarily stored. The light source S is not particularly limited, and may be any light source. As described later, in order to obtain image data of a spectral reflectance distribution from a multiband image, a light source having a known spectral intensity distribution is used. Preferably, there is.
Also, the multiband image has at least a band band of 8
It is divided into bands or more.

The multi-band image recording section 14 is a memory section for temporarily storing image data of a multi-band image composed of a plurality of images obtained by the multi-band camera 12,
Immediately after all the band images have been captured, the image data of the multi-band image is sent to the spectral reflectance distribution estimating unit 16.

The spectral reflectance distribution estimating section 16 is a section for estimating the spectral reflectance distribution of the object O from the obtained multi-band image, The band image is provided as reference image data, and the spectral reflectance is estimated by dividing the image data of each band image of the multiband image of the subject O by the reference image data value of the corresponding band image of the white plate. can do. Also, the spectral reflectance 16 is
A gray patch whose spectral reflectance is known, for example, six gray patches of a Macbeth chart, is used as reference image data, and a one-dimensional lookup table is created in advance in which the reference image data values correspond to the known spectral reflectance values. And this one
The spectral reflectance of the subject O may be estimated using a dimensional lookup table. The image data of the estimated spectral reflectance is sent to the image data synthesizing unit 24.

On the other hand, a light source spectral distribution database unit (hereinafter, referred to as a light source spectral distribution DB unit) 18 is a database unit having a plurality of spectral intensity distributions of the illumination light source. In the light source spectral distribution DB unit 18, the light sources are classified into indoor light sources and outdoor light sources, and the indoor light sources have, for example, uniform spectral intensity distributions of a white fluorescent lamp and an incandescent lamp. On the other hand, outdoor light source is configured with a spectral intensity distribution corresponding to the color temperature of the light sources, for example, the color temperature is the case of 6500K (Kelvin), corresponding to the spectral intensity distribution of light D ₆₅ of CIE standard . Further, as will be described later, the light source spectral distribution DB unit 18 is formed into a database such that a reference function T for determining a color temperature with respect to a time zone is uniquely determined according to a place name of a place designated as a search key. ing.

A background image database section (hereinafter referred to as a background image DB section) 20 is a database section provided with a plurality of multi-band background images to be synthesized as a background image of the subject O. The spectral intensity distribution of the light source is known. This multi-band background image may be recorded together with information on the shooting location and stored in a database. In this case, a background image that matches a location as a search key to be specified when selecting a spectral intensity distribution described later is used. Can be used to select. In addition,
It is preferable that the spectral component wavelength of the multi-band background image of the background image DB unit 20, that is, the wavelength representing each band is the same as the spectral component wavelength of the image data of the spectral reflectance distribution of the subject O. This is because, when the spectral component wavelengths are different, it is necessary to perform interpolation or extrapolation to make the spectral component wavelengths uniform.

The search unit 22 is a selection unit for selecting a desired light source spectral intensity distribution and a desired multiband background image from the light source spectral distribution DB unit 18 and the background image DB unit 20. To select the spectral intensity distribution of the light source, at least one of the location (p), season (s), and time zone (h) is designated as an inspection key. For an indoor light source, for example, an incandescent fluorescent lamp or an incandescent light bulb is selected. Select uniformly. On the other hand, in the case of an outdoor light source, as described later, the color temperature can be uniquely determined by specifying the location (p), season (s), and time zone (h) as a search key. More spectral intensity distribution can be determined. The obtained spectral intensity distribution is sent to the composite image generation unit 26.

With respect to the multi-band background image, a plurality of sample images of the background image are displayed as index images on a monitor (not shown) or the like, and the multi-band background image can be selected from these. Also, in conjunction with the key search location (p) at the time of selecting the light source spectral intensity distribution, a multiband background image corresponding to the location (p) may be selected. The obtained multiband background image is sent to the image data synthesizing unit 24.

The image data synthesizing section 24 obtains background image spectral reflectance data from the multiband background image selected by the search section 22, and obtains the image data and the subject spectral reflectance data estimated by the spectral reflectance estimating section 16. Are combined to obtain combined image spectral reflectance data. First, a band image of the multi-band image of the subject O is displayed as a monochrome image on a monitor (not shown). The shape of the subject O in the multi-band image is obtained by an operator's manual or automatically by a known subject edge detection method, and subject spectral reflectance data of pixels corresponding to this shape is extracted. Further, in order to determine the position of the subject O when combining with the background image, the band image of the multi-band background image is image-displayed on a monitor (not shown),
The position where the subject O is to be pasted is indicated by a mouse, keyboard, or the like (not shown). Thereafter, the component of the spectral intensity distribution of the photographing light that is known at the time of photographing the multi-band background image is removed from the multi-band background image to obtain background image spectral reflectance data. The background image spectral reflectance data of the background image area to be replaced with the subject O and replaced is replaced with the subject spectral reflectance data, and combined image spectral reflectance data is obtained. The obtained composite image spectral reflectance data is sent to the composite image generation unit 26.

The composite image generation unit 26 adds the composite image spectral reflectance data obtained by the image data
This is a part that generates image data of a composite image by multiplying the spectral intensity distribution of the light source obtained in step 2. That is, the image data of the multi-band image of the composite image is generated by multiplying the composite image spectral reflectance data by the spectral intensity distribution for each spectral component wavelength. The image data of the generated composite image is sent to the image data output unit 28.

The image data output section 28 converts the image data of the generated composite image into an output image signal suitable for the printer 30, and outputs the output image signal. The image data output unit 28 outputs the output image signal value (voltage value V _R ,
V _G and V _B ) are determined by a known method on a color chart with a known color chart, for example, CIEL ^* a
The values of L ^* , a ^* and b ^{* in} the ^* b ^* space are determined, and this L
^*, It comprises a three-dimensional look-up table created in advance by associating the a ^* and b ^* values and the output image signal value (voltage value V _R, V _G and V _B). On the other hand, since the composite image data is data having a value for each spectral component wavelength, the composite image data is multiplied by the value of the color-matching function of the XYZ color system for each spectral component wavelength, and the sum thereof is calculated. , Y and Z are obtained, and the three stimulus values X, Y and Z are obtained.
According to a known method, the L of the CIEL ^* a ^* b ^* space
The values of ^* , a ^*, and b ^* are calculated, and the colorimetric values are obtained. As a result, data can be converted into an output image signal using the three-dimensional lookup table. The obtained output image signal is sent to the printer 30 and can output a composite image.

Although the present apparatus 10 is configured as described above, the multi-band image of the subject O is not necessarily limited to the one captured and recorded by the multi-band camera 12, and the multi-band image of the subject O As long as the spectral intensity distribution of the illumination light source is known, there is no particular limitation,
For example, a multiband image of the subject O recorded in the database may be used.

Next, the operation of the multi-band image processing apparatus 10 will be described as an embodiment of the multi-band image processing method of the present invention. FIG. 2 shows an embodiment of the multi-band image processing method according to the present invention.

First, a multi-band image of the subject O is photographed and recorded by the multi-band camera 12, and image data values of the multi-band image are photographed in advance by the light source S for each spectral component wavelength representing a band band. The spectral reflectance of the object O is estimated by dividing the white plate whose spectral reflectance is known by the value of the reference image data (step 1).
00).

On the other hand, in order to select an illumination light source for the synthesized image, a light source spectral distribution DB unit 1 having a spectral intensity distribution of the light source is selected.
8 to select a desired light source spectral distribution (step 10).
2). The selection of the light source spectral distribution is performed as shown in FIG.
First, the location (p) is specified (step 102a).
The designation of a place designates a specific place name, indoor or the like. For example, items such as a specific place name and indoors of a place are displayed in a menu format on a monitor screen (not shown) in a hierarchical or enumerated manner, and an operator gives an instruction with a mouse or a keyboard. When the indoor light source is selected, a fixed light source such as an incandescent fluorescent lamp or an incandescent light bulb is displayed on a monitor screen (not shown) as a light source, and the operator specifies the light source alternatively (step 102b).

On the other hand, when a specific place name of a place is designated, the light source is an outdoor light source, and a reference function T relating to the time zone (h) of the color temperature corresponding to each place name is set (step 1).
02c). The reference function T is a function representing a temporal change in color temperature from sunrise to sunset, which is a standard for sunlight as an outdoor light source. The standard day is, for example, a spring equinox day or an autumn equinox day in which the time from sunrise to sunset is 12 hours. For example, when the location (p) is designated as Tokyo, a reference function T as shown on the left side of FIG. 4A is determined, and when it is designated as Kenya, the reference function T as shown on the right side of FIG. The reference function T is determined. Here, the time zone (h) that defines the reference function T is based on the approximate time zone in which the sun rises most, that is, at noon on the day of spring equinox or autumn equinox. The reference function T shown in FIGS. 4A to 4C is an elliptic function whose origin is the reference time zone (h). This is because before and after the reference time, the sun lowers the altitude and the color temperature decreases. However, FIG.
The reference function T shown in (a) to (c) is merely an example, and is not limited to this, and need not necessarily be elliptical.

Next, the operator designates the season (s), for example, spring, summer, autumn and winter (step 102).
d). The designation of the season (s) is because the time from sunrise to sunset changes depending on the season, and even in the same time zone (h), the altitude of the sun and the color temperature change depending on the season (s). Therefore, as shown in FIG. 4B, the reference function T determined in step 102c is
(Step 102) according to the season (s).
e). For example, when the season (s) is summer, the time from sunrise to sunset is long, so that it is expanded in the direction of the time axis h, and in winter, it is compressed. Specifying the season (s)
Moon instead of spring, summer, autumn and winter (January, February ...)
May be specified. Further, since the altitude of the sun rises and falls and the color temperature changes according to the season (s), the compression / expansion may be performed in the color temperature direction of the reference function T according to the season (s). Further, since the time zone in which the color temperature is maximum is not invariable throughout the year, the time zone in which the altitude of the sun is maximum depends on the origin of the time axis h of the reference function T according to the designation of the season (s). As such, the scaled reference function T may be translated in the direction of the time axis h.

Next, the operator designates a time zone (h), for example, a time zone such as 9:00 or 10:00 (step 102).
f). This is because the color temperature is determined by specifying the time zone (h). For example, if you specify 15:00 and time zone (h),
As shown in FIG. 4C, the color temperature in the corresponding time zone 15:00 on the expanded reference function T is determined, and the color temperature is determined (step 102g). The designation of the time zone (h) is not limited to the case where the time is designated, but may be a fixed time range such as 9 to 11 o'clock, and further designates a large time zone such as morning, noon and afternoon. It may be.

That is, in the case of an indoor light source, a white fluorescent lamp or an incandescent light bulb is alternatively determined as the light source, and the spectral intensity distribution is uniquely determined. On the other hand, in the case of an outdoor light source, the color temperature is determined from the place (p), the season (s), and the time zone (h), and the CIE standard light or auxiliary standard light corresponding to the color temperature, for example, FIG. As shown in FIG.
000K (Kelvin), the spectral intensity of the auxiliary standard illuminant D ₅₀ distribution, 5500K (Kelvin), the spectral intensity distribution of the auxiliary standard illuminant D ₅₅ is uniquely determined. Thus, the light source spectral distribution is determined (Step 102h). When the place is constant or the season is constant, or when the time zone is constant, the place (p), the season (s)
And at least one of the time zone (h) may be designated as the inspection key.

Next, a desired background image for synthesizing the subject O is selected from the background image DB unit 20, and a multi-band background image is selected (step 104). The multi-band background image may be selected by an operator's manual or in conjunction with a location (p) which is a search key when selecting the light source spectral distribution (step 102).

The background image spectral reflectance data obtained by dividing the known spectral intensity distribution from the multiband background image selected in step 104 and the subject spectral reflectance data obtained in step 100 are synthesized, and synthesized. Image data is obtained (step 106). The image data is synthesized by extracting the shape of the subject O, which is automatically detected by an operator's manual or by a known subject edge detection method, from the multiband image of the subject O. This is performed by extracting the spectral reflectance data and exchanging the spectral reflectance data of the background image of the desired area where the subject indicated by the background image is synthesized.

Next, the composite image data obtained in step 106 is multiplied by the spectral intensity distribution of the light source obtained in step 102 for each spectral component wavelength, thereby obtaining a composite image based on the light source obtained in step 102. Image data of the image is generated (step 108). At this time, the light intensity of the light source in the area of the background image may be set smaller than the light intensity of the light source in the area of the subject O. That is, the intensity of the light applied to the background image area and the intensity of the light source applied to the subject image area may be increased or decreased while maintaining the waveform of the spectral intensity distribution. The image data of the composite image is a multi-band image having an image data value for each spectral component wavelength and cannot be printed out as it is. Therefore, first, the image data is converted to CIEL ^* a ^* b which is a colorimetric value. ^* Space L ^* ,
The values are converted into values of a ^* and b ^* (step 110). In this conversion, first, image data of a composite image having a value for each spectral component wavelength is multiplied by the value of a color matching function of the XYZ color system for each corresponding spectral component wavelength, and the sum is calculated. X, Y, and Z are obtained, and the tristimulus values X, Y, and Z are used to calculate the values of L ^* , a ^*, and b ^{* in} the CIEL ^* a ^* b ^* space by a known method.

The data of the colorimetric values of the obtained composite image is represented by L ^* , a of a known color chart suitable for the printer 30.
^The output image signal is converted into an output image signal using a three-dimensional lookup table created in advance based on the values of ^* and b ^* and the output image signal (step 112). Thereafter, the composite image is output to the printer 30 and printed out (step 114).

As described above, when the subject O and the background image are combined, the subject O and the background image are combined and illuminated by the light source having the same spectral intensity. Thus, a synthesized image with a sense of reality can be obtained.

Although the multi-band image processing apparatus and processing method of the present invention have been described in detail above, the present invention is not limited to the above-described embodiment, and various improvements and modifications can be made without departing from the gist of the present invention. It goes without saying that it may be performed.

[0036]

As described above in detail, according to the present invention, a composite image is created by combining image data of the spectral reflectance of a subject and image data of the spectral reflectance distribution of a background image. Colorimetrically accurately reproduces the subject's image that matches the atmosphere at the time of the actual shooting, and the subject image fuses seamlessly with the background image to obtain a fully realistic composite image Can be. Also, the spectral intensity distribution of the light source when obtaining a composite image can be uniquely determined by specifying a location, a season, and a time zone as an inspection key, for example, when an outdoor light source is selected. In addition, a light source suitable for the background image can be selected. Further, since a composite image can be generated by changing the light source intensity between the subject image and the background image, a composite image having a more realistic feeling can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram illustrating an example of a multiband image processing device according to the present invention.

FIG. 2 is a flowchart illustrating an example of the flow of a multi-band image processing method according to the present invention.

FIG. 3 is a flowchart illustrating details of selection of a light source spectral distribution in the flow illustrated in FIG. 2;

FIGS. 4A to 4D are explanatory diagrams illustrating a light source spectral distribution.

[Explanation of symbols]

 Reference Signs List 10 Multi-band image processing device 12 Multi-band camera 14 Multi-band image recording unit 16 Spectral reflectance distribution estimating unit 18 Light source spectral distribution database unit 20 Background image database unit 22 Search unit 24 Image data synthesizing unit 26 Synthetic image generating unit 28 Image data Output unit 30 Printer

Claims (5)

[Claims] 1. A multi-band image processing apparatus for processing a multi-band image composed of a plurality of images obtained by dividing a photographing wavelength band into a plurality of band bands when photographing a subject. A light source spectral distribution database unit including a plurality of light source spectral intensity distributions; a background image database unit including a plurality of multi-band background images having a known spectral intensity distribution of the light source at the time of shooting to be combined as a subject background image; Selecting means for selecting a desired light source spectral intensity distribution and a desired multiband background image from the distribution database section and the background image database section; and a background image spectral reflection obtained from the multiband background image selected by the selecting means. Rate data and subject spectral reflectance data obtained from a multiband image of the subject. An image data combining unit that obtains combined image spectral reflectance data; and a combined image using the combined image spectral reflectance data obtained by the image data combining processing unit and the spectral intensity distribution selected by the selection unit. A composite image generation unit that obtains image data of the image data, and an image data output unit that converts image data of the composite image obtained by the composite image generation unit into a predetermined output image signal and outputs the signal. Image processing device for band images. 2. A multi-band image processing method for processing a multi-band image including a plurality of images obtained by dividing a photographing wavelength band into a plurality of band bands when photographing a subject. A spectral intensity distribution of a desired illumination light source is selected from a light source spectral intensity distribution database unit having a plurality of light intensity spectral intensity distributions. From a background image database unit having a plurality of multi-band background images with known light intensity spectral intensity distributions at the time of shooting. A desired multi-band background image to be synthesized as a subject background image is selected, and the background image spectral reflectance data obtained from the selected multi-band background image and the subject spectral reflectance obtained from the multi-band image of the subject Data and the combined image spectral reflectance data to obtain combined image spectral reflectance data and the spectral intensity distribution. And obtaining image data of the composite image, converting the image data of the composite image into a predetermined output image signal, and outputting the output signal. 3. The light source spectral distribution database unit includes a spectral intensity distribution corresponding to a color temperature of an outdoor light source. When selecting an outdoor light source as a desired illumination light source, at least one of a place, a season, and a time zone. 3. The image processing method for a multi-band image according to claim 2, wherein a color temperature of the light source is determined from the search key, and a spectral intensity distribution of the light source is determined from the color temperature. 4. When selecting a desired spectral intensity distribution of a light source for illumination, a reference function that defines a color temperature of an outdoor light source with respect to a time zone is set according to the specified location. Compressing / expanding at a ratio according to the designated season, and determining a color temperature in the designated time zone from the compressed / expanded reference function, thereby determining a color temperature from the search key. 3. The image processing method for a multi-band image according to 3. 5. The image data of the composite image is obtained by changing the intensity of a light source used for a background image area and a subject image area in the image data of the composite image when obtaining the image data of the composite image. An image processing method for a multiband image according to any one of the above.