JP2007165995A - Image generating device, image generating method, and image generating program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image generating device, an image generating method, and an image generating program capable of achieving accurate color reproduction and reducing a calculation cost even when a light source environment is varied. <P>SOLUTION: A light source environment measurement device 10 uses a multi-spectral moving picture camera to photographs light sources and quantizes camera signals by a weighted function corresponding to each light source. A light source spectrum estimate section 12 calculates a light source spectrum estimate matrix from camera sensitivity. A rendering processing section 15 uses information (a refractive index of an object and an intensity distribution function) of each light source of an object/scene model storage section 14 to calculate information associated with the reflection intensity of the surface of the object by each light source, a display signal generation section 18 generates a display signal displayed on a display device 19 on the basis of the camera signal of a light source spectrum information storage section 13, the light source spectrum estimate matrix, a rendering result of the rendering result storage section 16, and display device spectrum characteristic data of a display device spectrum characteristic data storage section 17. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image generation apparatus for faithful color reproduction and texture reproduction, and more particularly to an image generation apparatus, an image generation method, and an image generation program that efficiently generate an image even when a light source environment changes.

  Since the appearance of an object changes greatly with fluctuations in the light source environment, in order to faithfully reproduce the real world (simulation), the real world light source environment is acquired, and images / videos are displayed based on the light source environment. It is important to generate. However, when an image is generated with respect to a change in the light source environment, there is a problem that the calculation cost is increased in consideration of mutual reflection and the like.

  In order to solve such a problem, a technique for fusing a virtual object to the real world without a sense of incongruity has been proposed (for example, see Patent Document 1). In this prior art, the light source environment is represented by a large number of light sources on a hemisphere, and an original image is generated for each light source (the RGB value of the light source is assumed to be unit luminance) in consideration of mutual reflection that requires a calculation cost in advance. When generating an image in consideration of all light source environments, a linear sum of these is taken.

  The variation of the light source is dealt with by correcting the RGB value of the original image generated for each light source based on the RGB value of the light source environment actually measured (or virtually set). If this concept is used, it is not necessary to redo image generation in consideration of mutual reflection which requires a calculation cost, and an image corresponding to light source fluctuation can be generated in a short time. As a method for measuring the lighting environment, there is a method using an RGB camera with a fisheye lens.

  The reproduction of the virtual object according to the above-described prior art will be described in more detail. In the prior art, color reproduction based on RGB values (or tristimulus values) is performed in the process of light source, rendering, display signal generation, and all color reproduction processes. In order to perform accurate color reproduction with respect to fluctuations in the light source spectrum, in principle, spectral color reproduction is required, and the level of color reproduction is different. Hereinafter, a description will be given with reference to FIG.

The light source environment at each time t is photographed by the RGB camera of the light source environment measuring apparatus 1. In order to represent the light source environment with a plurality of light sources (j = 1,..., N), RGB values L ′ _j quantized by weighted functions corresponding to the respective light sources are acquired from the camera signal, and the light source RGB value storage unit 2 Save to.

The object / scene model storage unit 3 includes object / scene model data, RGB-based reflectance information R (θ _i , φ _i , θ _e , φ _e , λ) (λ = R, G, B) of the subject, The intensity distribution function D _j (θ _i , φ _i ) of each light source and the RGB values (L _j (R), L _j (G), L _j (B) of each light source L _j (j = 1,..., N) )) Is stored. The intensity distribution function is constant with respect to time and wavelength.

In the rendering processing unit 4, rendering processing is performed using information in the object / scene model storage unit 3. Under a certain light source j, brightness P _j (λ) at a wavelength λ (λ = R, G, B) (corresponding to R, G, B) of an object surface viewed from the viewpoint (θ _e , φ _e ) {Λ = R, G, B}

  And is stored in the rendering storage unit 5.

In the display signal generation unit 6, the total RGB value is obtained from the RGB value L ′ _j of each light source stored in the light source RGB value storage unit 2 and the rendering result of the rendering storage unit 5.

  Is calculated and output to the display device 7. However, the calculated RGB values are relative values, and an appropriate coefficient is multiplied so that the maximum value becomes the maximum value of the display signal.

  As another conventional technique, an accurate color reproduction technique based on a spectrum is shown, and faithful color reproduction is possible by generating an image using a light source spectrum and a spectral reflectance of a subject. A CG creation method has been proposed (Non-Patent Document 1).

As another conventional technique, a method of generating an image in a camera signal space using a spectral reflectance estimation technique using a multispectral camera is known (Non-Patent Document 2). This technology makes it possible to reproduce colors based on the spectrum with a realistic number of dimensions, and only needs to generate an image that takes into account the mutual reflection, which is computationally expensive. Even if there is a change in the illumination spectrum, recalculation is possible. Is unnecessary.
JP 2002-117413 A Yinlong Sun, F. David Fracchia, Mark S. Drew, Thomas W. Calvert “A spectarlly based framework for realistic image synthesis”, The Visual Computer pp.429-444 (2001), Published online 2 October 2001. M.Tsuchida, H.Arai, M.Nishiko, Y.Sakaguchi, T.Uchiyama, H.Haneishi, M.Yamaguchi, N.Ohyama ”Development of BRDF and BTF Measurement and Computer-aided Design Systems Based on Multispectral Imaging” ” AIC Color 05 "Proceedings pp.129-132 (2005).

  The prior arts disclosed in Patent Document 1 described above are all based on the RGB values of the light source when generating and correcting the original image. In order to reproduce an accurate color in a situation where the light source changes, it is necessary to consider the light source spectrum and the spectral reflectance information of the subject. However, the RGB value (or XYZ tristimulus value) of the light source is the RGB value of the completely diffuse reflecting object, and the change in the RGB value of an arbitrary reflectance object cannot be accurately predicted from this RGB value.

  At the time of CG image generation, generally, after considering illumination, image generation is performed in a space of XYZ or RGB tristimulus values. However, when the tristimulus value is reached, the light source spectrum and the spectral reflectance of the subject cannot be separated, and it is difficult to generate a faithful image from the tristimulus image corresponding to the variation of the light source spectrum. It is.

  Further, in the conventional technique according to Non-Patent Document 1, an accurate color reproduction technique based on a spectrum is shown. However, since an image is generated in a spectrum space, if the number of dimensions representing the spectrum is large, the calculation amount is enormous There is a problem.

  In addition, according to the prior art according to Non-Patent Document 2, color reproduction based on a spectrum can be performed with a realistic number of dimensions, and image generation taking into account mutual reflection, which requires calculation cost, may be performed once. If there is, recalculation is unnecessary. However, the light source environment that spreads on the hemisphere and fluctuates in real time cannot be obtained in a spectral manner, and it is difficult to perform color reproduction and image generation using it.

  The present invention has been made in view of such circumstances, and an object thereof is to realize accurate color reproduction even when the light source environment fluctuates, and to reduce calculation costs. Another object of the present invention is to provide an image generation apparatus, an image generation method, and an image generation program method.

  In order to solve the above-described problems, the present invention provides an image generation apparatus that generates a target image for displaying a subject whose light source environment changes on the display apparatus, the display apparatus including a plurality of light source environments. Generated by the light source environment measuring means for measuring the time series information of the spectrum corresponding to the light source representing the light source, the light source spectrum estimating means for estimating the light source spectrum estimation matrix based on the sensitivity characteristic relating to the predetermined measurement, and the light source environment measuring means Light source spectrum information storage means for storing the time-series information of the spectrum obtained and the light source spectrum estimation matrix estimated by the light source spectrum estimation means, and at least stored in the light source spectrum information storage means. Based on the time series information and the light source spectrum estimation matrix, the light source environment under the light source environment Generate the target image of Utsushitai, characterized in that the object image generated and a display image generating means for outputting to the display device.

  The present invention, in the above invention, comprises rendering processing means for calculating reflection information relating to the reflection intensity of the subject surface under a certain light source using the reflectance of the subject and the intensity distribution function of each light source, and the display image generation The means is a light source based on the time-series information of the spectrum, the light source spectrum estimation matrix, and the information on the reflection intensity calculated by the rendering processing means stored in the light source spectrum information storage means. The target image of the subject under the environment is generated.

  According to the present invention, in the above invention, the light source environment measuring unit includes a multispectral video camera that captures a spectrum corresponding to a light source representing a plurality of light source environments at each time, and the light source spectrum estimating unit includes the multispectral The light source spectrum estimation matrix is estimated based on the sensitivity characteristic of the moving image camera.

  In order to solve the above-described problem, the present invention is an image generation method in an image generation apparatus that generates a target image that is connected to a display device and displays a subject whose light source environment changes on the display device. Measuring spectrum time series information corresponding to light sources representing a plurality of light source environments, estimating a light source spectrum estimation matrix based on sensitivity characteristics relating to predetermined measurement, and time series information of the spectrum, , Storing the estimated light source spectrum estimation matrix, and generating the target image of the subject under a certain light source environment based on at least the time-series information of the spectrum and the light source spectrum estimation matrix And a step of outputting the generated target image to the display device. It is a generation method.

  In order to solve the above-described problem, the present invention provides a computer of an image generation apparatus that generates a target image that is connected to a display device and displays a subject whose light source environment changes on the display device. Measuring time series information of a spectrum corresponding to a light source representing a light source environment; estimating a light source spectrum estimation matrix based on a sensitivity characteristic related to a predetermined measurement; time series information of the spectrum; Storing the light source spectrum estimation matrix, generating the target image of the subject under a certain light source environment based on at least the time-series information of the spectrum and the light source spectrum estimation matrix, and generating An image generation program for executing the step of outputting the target image to the display device That.

  According to the present invention, the image generation apparatus measures time-series information of a spectrum corresponding to a light source representing a plurality of light source environments, estimates a light source spectrum estimation matrix based on a sensitivity characteristic related to a predetermined measurement, The time series information and the light source spectrum estimation matrix are stored in the light source spectrum information storage means. Further, a target image of a subject under a certain light source environment is generated based on at least the stored time-series information of the spectrum and the light source spectrum estimation matrix, and the generated target image is output to the display device. As a result, even when the light source environment fluctuates, accurate color reproduction can be realized, and when performing image generation processing, spectral information is maintained, so that all processing is not repeated. This can be implemented and the calculation cost can be reduced.

  Further, according to the present invention, the image generation apparatus calculates and stores the reflection information related to the reflection intensity of the subject surface under a certain light source for each light source using the reflectance of the subject and the intensity distribution function of each light source. The target image of the subject under a certain light source environment is generated based on the time-series information of the current spectrum, the light source spectrum estimation matrix, and the information on the reflection intensity. As a result, even when the light source environment fluctuates, accurate color reproduction can be realized, and when performing image generation processing, spectral information is maintained, so that all processing is not repeated. This can be implemented and the calculation cost can be reduced.

  Further, according to the present invention, the image generation device captures a spectrum corresponding to a light source representing a plurality of light source environments at each time with a multispectral video camera, and the light source spectrum is based on sensitivity characteristics of the multispectral video camera. By estimating the estimation matrix, time-series information of the spectrum is measured. As a result, even when the light source environment fluctuates, accurate color reproduction can be realized, and when performing image generation processing, spectral information is maintained, so that all processing is not repeated. This can be implemented and the calculation cost can be reduced.

Hereinafter, an image forming apparatus according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram illustrating a configuration of an image forming apparatus according to an embodiment of the present invention. In the figure, a light source environment measuring apparatus 10 is a multi-spectrum video camera (for example, a camera equipped with a fisheye lens) that generates time-series information (referred to as a time-series light source environment) of a spectrum corresponding to each light source representing the light source environment. I have. The light source environment measurement apparatus 10 quantizes a camera signal obtained by photographing a light source with a multispectral video camera using a weighted function corresponding to each light source and supplies the quantized function to the light source spectrum information storage unit 13.

  The light source measurement device spectral characteristic data storage unit 11 stores camera sensitivity (spectral characteristic data) of the multispectral video camera of the light source environment measurement device 10. The light source spectrum estimation unit 12 calculates a light source spectrum estimation matrix from the camera sensitivity stored in the light source measurement device spectral characteristic data storage unit 11 and supplies it to the light source spectrum information storage unit 13. The light source spectrum information storage unit 13 stores the quantized camera signal from the light source environment measurement device 10 and the light source spectrum estimation matrix estimated by the light source spectrum estimation unit 12.

  The object / scene model (BRDF) storage unit 14 stores object / scene model data, the reflectance of the subject (bidirectional reflection distribution function), and the intensity distribution function of each light source. Here, the reflectance of the subject is obtained in advance by a multispectral camera that measures the reflectance. Hereinafter, the multispectral camera for measuring the reflectance is referred to as a reflectance measuring multispectral camera. The rendering processing unit 15 uses the information stored in the object / scene model storage unit 14 (subject reflectance (bidirectional reflection distribution function), intensity distribution function of each light source) for each light source. Information on the reflection intensity of a certain object surface as viewed from the viewpoint is calculated (rendering process). The rendering result storage unit 16 stores information (rendering result) on the reflection intensity of the object surface for each light source calculated by the rendering processing unit 15.

  The display device spectral characteristic data storage unit 17 stores spectral characteristic data indicating spectral characteristics of the display device 19 to be described later. The display signal generation unit 18 stores the camera signal and the light source spectrum estimation matrix stored in the light source spectrum information storage unit 13, the rendering result stored in the rendering result storage unit 16, and the display device spectral characteristic data storage unit 17. A display signal is generated from the stored display device spectral characteristic data and output to the display device 19. For example, an RGB display device, a multi-primary color display device, or a printer is applied to the display device 19, and the image data output from the display signal generation unit 18 is displayed or printed.

  In this embodiment, the conventional technique performs color reproduction based on RGB values (or tristimulus values) in the process of light source, rendering, display signal generation, and all color reproduction processes. The difference is that this is a method for realizing spectral color reproduction (spectral color reproduction). The principle of the image forming method according to the present embodiment will be described below.

Here, FIG. 2 shows a case in which illumination light is incident on a certain point on the subject surface from an angle (θ _i , φ _i ) and reflected light reflected from the point at the angle (θ _e , φ _e ) is photographed. It is the figure which showed the positional relationship. As illustrated, the light source direction is represented by (θ _i , φ _i ) and the line-of-sight direction is represented by (θ _e , φ _e ) when viewed from the one point. The light source environment is represented by a number of light sources (point light source, surface light source) (j = 1,..., N) on the hemisphere.

  First, an original image of a virtual object is generated separately for each light source under a light source (E light source) having the same intensity at all wavelengths. Details of generation (formula etc.) will be described later.

  In order to spectrally acquire a changing light source environment, a light source is photographed by a multispectral video camera provided in the light source environment measuring apparatus 10, and a spectrum time series corresponding to each light source (j) representing the light source environment is obtained. Information (this is called a time-series light source environment) is generated. For example, if a fisheye lens or the like is used, hemispherical light source information can be acquired. When setting that the light source position is at infinity is difficult, measurement is performed with a plurality of cameras, or the distance to the light source is obtained by another means to obtain the three-dimensional spatial position information of the light source. . In the following description, it is assumed that the light source position is at infinity.

When the time axis is represented as t, the sensitivity of the multispectral video camera provided in the light source environment measuring apparatus 10 is S ^L , the spectrum of each light source is L _j (t), and the camera signal is v ^L _j (t).

It can be expressed. The spectral spectrum L _j (t) of the light source can be estimated from the camera signal v ^L _j (t) by Wiener estimation or the like,

It can be expressed as. Here, ^GL is a spectrum estimation matrix. It is also possible to virtually generate a time-series light source environment regardless of measurement. Assuming that the spectral spectrum is constant in each light source (j), the illumination L _j is represented by a spectral distribution function L (λ) and an intensity function D (θ _i , φ _i ) determined from the shape and illumination characteristics of the illumination.

It can be expressed as Here, θ _i and φ _i represent the incident angle of light.

  Next, faithful color reproduction is realized reflecting the spectral light source environment, and when performing image generation processing, spectral information is maintained and all processing can be performed without re-execution. How to do.

  Here, it is assumed that the bidirectional reflection characteristic is measured by a multispectral camera for reflectance measurement. It should be noted that the same processing can be performed by converting the spectrum information measured by the spectrometer into a multispectral signal using the virtual camera sensitivity. When considering BTF (Bidirectional Texture Function), the same processing can be performed by adding image coordinate information (x, y) to the reflectance.

Consider rendering at a point on the object surface at a certain time. It is assumed that the characteristic of the object surface can be expressed as R (θ _i , φ _i , θ _e , φ _e ) by the bidirectional reflection distribution function BRDF (Bidirectional Reflectance Distribution Function). Here, θ _e and φ _e are line-of-sight directions. When the object surface is irradiated with illumination having a spectral characteristic of L ₀ and taken with a multispectral camera for measuring the reflectance of the spectral characteristic S, the camera signal v is

  It can be expressed. The reflectance is estimated from the camera signal v by Wiener estimation.

  Can be estimated. Here, G is a reflectance estimation matrix. The spectrum s obtained from the entire light source environment can be expressed by the following equation using the illuminance obtained from each light source and the bidirectional reflection distribution function BRDF.

  For this reason, virtual object image generation under a light source environment in the real world can be realized by decomposing each light source into a set of point light sources and then capturing and storing an object brightness image for each point light source. Here, according to Equation (5)

It can be expressed. Furthermore, since R (θ _i , φ _i , θ _e , φ _e ) = Gv (θ _i , φ _i , θ _e , φ _e ) according to Equation (7),

  It can be expressed. K represents the number of bands of the multispectral camera for reflectance measurement. Since G (λ, k) is a constant for integration,

  And can be transformed. here,

  And considering over all wavelengths λ,

It can be expressed. L _j in Equation (13) represents a spectral spectrum with a diagonal matrix.

From the above, it is possible to perform spectrum calculation based on the light source environment including the spectrum change if the calculation corresponding to Equation (12) is performed once by the intensity function D _j (θ _i , φ _i ) of each light source environment. Become. An original image of each light source is obtained by performing Expression (12) on all image coordinates. In the actual calculation, each pixel is converted from the original image into a tristimulus value or a signal value of the display device through spectrum calculation, and it is necessary to store the spectrum calculation result in a variable or output it as an image. There is no.

  Next, the operation of the above-described embodiment will be described. FIG. 3 is a flowchart for explaining the operation of the image forming apparatus according to the present embodiment. In this embodiment, both the light source and the reflectance of the object model are represented by the band value of the multispectral camera for reflectance measurement and the spectrum estimation matrix (specifically, the spectral radiance estimation matrix and the spectral reflectance estimation matrix). Spectral color reproduction is realized using spectral information. Hereinafter, a description will be given with reference to FIG.

The light source environment measuring device 10 captures the light source environment at each time t with a multispectral video camera provided in the light source environment measuring device 10 (S1). In order to represent the light source environment with a plurality of light sources (j = 1,..., N), the camera signal v ^L _j quantized by the weighted function corresponding to each light source is obtained from the camera signal, and the light source spectrum information storage unit 13 (S2).

In the light source spectrum estimating unit 12 stores the camera speed S ^L of the light source measuring device spectral characteristic data storage unit 11, calculates the source spectrum estimation matrix G ^L to the light source spectrum information storage unit 13 (S3). Examples of calculation formulas are as follows.

  Here, ρ = 0.999 and M is the number of dimensions expressing the spectrum.

The object / scene model storage unit 14 stores object / scene model data, subject reflectance (bidirectional reflection distribution function) R (θ _i , φ _i , θ _e , φ _e , λ), and intensity distribution of each light source. It is assumed that the function D _j (θ _i , φ _i ) is stored. The reflectance R (θ _i , φ _i , θ _e , φ _e , λ) (λ = 1,..., M) is the multispectral camera signal v (θ _i , φ _i , θ _e , φ _e , k) ( k = 1,..., K) and the reflectance estimation matrix G,

  It can be expressed as Here, k (k = 1,..., K) represents a band of the multispectral camera for reflectance measurement, and K represents the number of bands. In the matrix (vector) representation, k, λ, etc. representing each element are not shown.

The rendering processing unit 15 performs rendering processing using information in the object / scene model storage unit 14. As information on the reflection intensity of a certain object surface viewed from the viewpoint (θ _e , φ _e ) under a certain light source j,

  Is calculated. This is calculated for each light source and stored in the rendering result storage unit 16 (S4).

In the display signal generation unit 18, the camera signal v ^L (t) and the light source spectrum estimation matrix G ^L stored in the light source spectrum information storage unit 13, the rendering result v ′ (k, j) in the rendering result storage unit 16, and the display device Using the display device spectral characteristic data in the spectral characteristic data storage unit 17 as an input, a display signal α is generated and output to the display device 19 (S5). The spectral signal s is

  Can be calculated. An expression for calculating the display signal α from this spectrum signal is shown below. The following example is considered as display device spectral characteristic data. Assume that the display is a display device in which the bias can be ignored and the tone curve is linear, the primary color spectrum is represented by P (M × N matrix), and the color matching function CMF is represented by X (3 × M matrix). Here, N is the number of primary colors. At this time, according to the tristimulus value matching condition,

  P is spectral characteristic data of the display device 19 stored in the display device spectral characteristic data storage unit 17, and X is a value preset in the display signal generation unit 18.

  It is. When the number of primary colors N is 3,

  Thus, the display signal α can be calculated. Several algorithms have been proposed when the number of primary colors is 4 or more. The display signal α is calculated from these series of equations. In actual calculation, it is not necessary to calculate the spectrum s as an intermediate product, and substitution is performed on the mathematical formula. Therefore, the number of dimensions in the calculation is determined by the number of dimensions of the input signal and the display signal.

  According to the above-described embodiment, even when the light source environment fluctuates, accurate color reproduction can be realized, and spectrum information is maintained when performing image generation processing. It can be carried out without re-processing.

  In the above-described embodiment, a series of processes by the light source spectrum estimation unit 12, the rendering processing unit 15, the display signal generation unit 18, and the like described above are stored in a computer-readable recording medium in the form of a program. The above processing is performed by the computer reading and executing this program. That is, each processing means and processing unit in the light source spectrum estimation unit 12, the rendering processing unit 15, the display signal generation unit 18 and the like are read by the central processing unit such as a CPU into the main storage device such as a ROM and a RAM. This is realized by executing information processing / arithmetic processing.

  Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

1 is a block diagram illustrating a configuration of an image forming apparatus according to an embodiment of the present invention. It is a conceptual diagram for demonstrating the positional relationship of the to-be-photographed object, illumination light source, and imaging device in this embodiment. It is a flowchart for demonstrating the image generation process by this embodiment. It is a block diagram which shows the structure of the image forming apparatus by a prior art.

Explanation of symbols

10 Light source environment measuring device (light source environment measuring means, multispectral video camera)
11 light source measurement device spectral characteristic data storage unit 12 light source spectrum estimation unit (light source spectrum estimation means)
13 Light source spectrum information storage unit (light source spectrum information storage means)
14 Object / Scene Model Storage Unit 15 Rendering Processing Unit (Rendering Processing Unit)
16 Rendering Result Storage Unit 17 Display Device Spectral Characteristic Data Storage Unit 18 Display Signal Generation Unit (Display Image Generation Unit)
19 Display device

Claims (5)

An image generation device that generates a target image that is connected to a display device and displays a subject whose light source environment changes on the display device,
A light source environment measuring means for measuring time-series information of a spectrum corresponding to a light source representing a plurality of light source environments;
Light source spectrum estimation means for estimating a light source spectrum estimation matrix based on a sensitivity characteristic relating to a predetermined measurement;
Light source spectrum information storage means for storing time-series information of the spectrum generated by the light source environment measurement means and a light source spectrum estimation matrix estimated by the light source spectrum estimation means;
Based on at least the time-series information of the spectrum stored in the light source spectrum information storage unit and the light source spectrum estimation matrix, the target image of the subject under a certain light source environment is generated and the generated An image generation device comprising: a display image generation unit that outputs a target image to the display device. Rendering processing means for calculating reflection information on the reflection intensity of the subject surface under a certain light source using the reflectance of the subject and the intensity distribution function of each light source,
The display image generation means includes
Based on the time-series information of the spectrum, the light source spectrum estimation matrix, and the information on the reflection intensity calculated by the rendering processing means stored in the light source spectrum information storage means, under a certain light source environment The image generation apparatus according to claim 1, wherein the target image of the subject is generated. The light source environment measuring means includes
A multi-spectrum video camera that captures spectra corresponding to light sources representing multiple light source environments at each time;
The image generation apparatus according to claim 1, wherein the light source spectrum estimation unit estimates a light source spectrum estimation matrix based on sensitivity characteristics of the multispectral video camera. An image generation method in an image generation apparatus that generates a target image that is connected to a display device and displays a subject whose light source environment changes on the display device,
Measuring spectral time series information corresponding to light sources representing multiple light source environments;
Estimating a light source spectrum estimation matrix based on a sensitivity characteristic relating to a predetermined measurement;
Storing the time-series information of the spectrum and the estimated light source spectrum estimation matrix;
Generating the target image of the subject under a certain light source environment based on at least the time-series information of the spectrum and the light source spectrum estimation matrix;
And a step of outputting the generated target image to the display device. A computer connected to a display device for generating a target image for displaying a subject whose light source environment changes on the display device;
Measuring spectral time series information corresponding to light sources representing multiple light source environments;
Estimating a light source spectrum estimation matrix based on a sensitivity characteristic relating to a predetermined measurement;
Storing the time-series information of the spectrum and the estimated light source spectrum estimation matrix;
Generating the target image of the subject under a certain light source environment based on at least the time-series information of the spectrum and the light source spectrum estimation matrix;
An image generation program for executing the step of outputting the generated target image to the display device.

Images