JP2733412B2 - Reflection model creation method and image data manipulation device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a reflection model for obtaining a reflection model representing image data of a part of an object when performing processing for changing the color, surface shape, and properties of the object included in the image. How to create. Further, the present invention relates to an image data manipulation device using the reflection model creating method.

When designing posters and merchandise with a computer-based design system and when creating computer graphics, a color image obtained by reading an original such as a photograph or an object portion included in the created color image is colored. And size and shape,
In many cases, a process of combining with another image is performed. Further, it is also necessary to change the glossiness and texture of the surface of the object in order to simulate a case where the material is changed. The same processing is also widely used in desktop publishing systems, printing original plate creation systems, and the like.

[0003] In such a computer system, it is necessary to obtain a high-quality image which does not cause unnaturalness to human vision by performing processing such as change of color, shape and surface properties with high accuracy. At the same time, there is a need to reduce the time required for these processes. There is also a demand for a simple and easy-to-use computer-to-human interface (human interface).

[0004]

2. Description of the Related Art As a method of changing the color of an object in an image, there is a method of changing the hue value of the color of each pixel of the object according to the designation of a color by an operator. In this case, the image data of each pixel is converted in advance to a color system using hue, saturation, and brightness as indices, and the hue value is changed.
The image data may be converted to the RGB color system and used for display on a display device or the like.

According to this method, the color of an object can be changed while preserving the surface properties such as the glossiness of the object. In addition, a color palette for specifying a target color is displayed on the display device, and the operator can easily specify the target color by a position on the screen where a desired color is displayed with a mouse or a light pen. , A simple human interface has been realized.

There is another method of changing the color of an object using a simple reflection model. On the other hand, as a method of changing the property of the surface of the object, there is a method in which an image of an object having a different material is input and held in advance, and is replaced with a part of the object cut out from the image to be processed.

[0007]

By the way, as described above, the method of changing the hue value of an object aims at changing the color of the object, and it is not possible to change the properties of the surface of the object. could not. Also, if you change the saturation or brightness to fine-tune the color of an object, shadows may become whitish or glossy, and the processed image may become unnatural. there were.

Similarly, even when a method using a simple reflection model is applied, the properties of the surface of the object cannot be changed. On the other hand, if a method of inputting an image of an object having a different material or color and replacing it with an image of a portion of the object is used, both the color and the surface property of the object can be changed. However, in this case, it is necessary to prepare an image of the object having the changed color or material in advance, and the preparation is time-consuming. In addition, the selection range of the characteristics such as the material and the color is limited to the characteristics of the object in the prepared image.

As a technique for solving these problems and obtaining an image perceived as a natural image by human vision while freely changing the color and surface properties of an object, the present applicant has already proposed And Japanese Patent Application No. 3-195326, entitled "Color Image Control Method".

In this technique, image data of a part of an object is represented using a reflection model that takes into account the color of the object itself (hereinafter, referred to as the object color) and the colors of illumination light and secondary reflected light from the surroundings. , Illumination light, and secondary reflected light, and the coefficients corresponding to the vectors representing these colors can be independently changed.

In the above-described reflection model, the image data (R ₀ , G ₀ , B ₀ ) is based on the object color vector (R _d , G
_d , B _d ), the illumination light vector (R _s , G _s , B _s ) and the secondary reflected light vector (R _a , G _a , B _a ), and the coefficients K ₁ , K ₂ corresponding to these vectors, respectively. , K _3, and is represented as an equation.

[0012]

(Equation 01)

Here, coefficients K ₁ ,
K ₂ and K ₃ are determined by the diffuse reflection characteristics, specular reflection characteristics, surface roughness, and observer's line of sight at each position of the object, and represent the shape of the object and the properties of the surface at each position. is there.

Therefore, in the equation, the object color vector, the illumination light vector, the secondary reflected light vector and the coefficient K
_By changing ₁ , K ₂ , and K ₃ independently, the color and surface properties of the object can be freely adjusted for objects with a uniform surface color, such as plastic shaped objects and painted objects. Can be obtained, image data of each pixel of the object portion can be obtained.

As described above, in the reflection model represented by the equation, since the coefficients K ₁ , K ₂ , and K ₃ are functions of the position, the coefficients K ₁ , K ₂ , and K ₃ are based on each vector estimated from the image data and the image data of each pixel. And the coefficients K ₁ , K ₂ ,
It is necessary to obtain the K _3.

Since the amount of arithmetic processing for obtaining these coefficients K ₁ , K ₂ , and K ₃ is enormous, it has been difficult to speed up the processing for changing the color and surface properties using the above-described technique. .

An object of the present invention is to provide a reflection model creating method capable of quickly obtaining a reflection model for changing the color and surface properties of an object with high accuracy. It is another object of the present invention to provide an image data manipulation device that performs a process of changing the color and surface properties of an object using the reflection model creation method.

[0018]

FIG. 1 is a diagram showing the principle of a reflection model creating method according to the present invention. According to the first aspect of the present invention, each component obtained by multiplying a vector representing the color of the object itself, the color of the illumination light, and the color of the secondary reflected light included in the image by a coefficient corresponding to each vector,
In a reflection model creating method for obtaining a reflection model representing image data representing a color of each pixel of an object portion, an illumination light vector indicating an illumination light color and a color of a secondary reflected light are indicated based on the image data. One of the secondary reflected light vectors and an object color vector indicating the color of the object itself are determined, and one of the illumination light vector and the secondary reflected light vector is determined using a vector corresponding to an ideal white light. Approximate to eliminate the component including the vector corresponding to the ideal white light from the expression representing the reflection model, to derive a system of binary equations regarding the object color vector and the illumination light vector or the secondary reflected light vector, It is characterized in that the binary simultaneous equations are solved to obtain coefficients corresponding to the object color vector, the illumination light vector, and the secondary reflected light vector, respectively.

FIG. 2 is a diagram showing the configuration of the image data manipulation device according to the second aspect. According to a second aspect of the present invention, there is provided an image storage means for storing image data representing a color of each pixel constituting an image, an object color vector indicating a color of an object included in the image based on the image data, and an illumination light. A reflection model generating means 102 for obtaining an illumination light vector indicating the color of the second light, a second reflected light vector indicating the color of the second reflected light, and a coefficient corresponding to each of these vectors, an object color vector, an illumination light vector, and a second Vector holding means 103 for holding a reflected light vector, and coefficient holding means 104 for holding a coefficient corresponding to each vector for each pixel
And image data calculation means 105 for calculating the image data of each pixel by multiplying the object color vector, the illumination light vector and the secondary reflection light vector by the corresponding coefficients, and sending the image data to the image storage means 101. In the image data manipulation device having an object color vector, an illumination light vector, a secondary reflection light vector, and a coefficient corresponding to each of these vectors in accordance with an input, a reflection model creation is performed. The means 102 generates a vector indicating ideal white light as a corresponding vector according to the input of the approximation instruction for specifying one of the illumination light vector and the secondary reflected light vector, and
3, an illumination light vector or a secondary reflected light vector and an object color vector are determined based on the image data stored in the image storage means 101 in accordance with the approximation instruction. According to the vector determining means 112 to be sent to the
The respective vectors and image data obtained by the vector determination means 112 are substituted into predetermined binary simultaneous equations relating to the illumination light vector or the secondary reflection light vector and the object color vector, and coefficients and approximation instructions corresponding to these vectors are used. It is characterized by comprising a coefficient calculating means 113 for obtaining a coefficient corresponding to the designated vector and sending it to the coefficient holding means 104.

[0020]

According to the first aspect of the present invention, the illumination light vector or the secondary reflection light vector is obtained from the expression representing the reflection model by utilizing that all the components of the vector representing the ideal white light have the same value. A coefficient corresponding to each of the object color vector, the illumination light vector, and the secondary reflected light vector is calculated by eliminating the included components and solving the obtained binary simultaneous equation. Here, it is clear that the processing for solving the binary simultaneous equation requires a smaller amount of processing than the processing for solving the ternary simultaneous equation obtained from the original reflection model. Therefore, by using this method, it is possible to reduce the time required for the process of calculating the coefficient corresponding to each vector described above, and to quickly perform a reflection model for changing the color and surface properties of an object with high accuracy. It is possible to obtain.

According to a second aspect of the present invention, based on the two vectors determined by the vector determination means in response to the approximation instruction, the coefficient calculation means 113 erases the component including the vector specified by the approximation instruction. The coefficients corresponding to the object color vector, the illuminating light vector, and the secondary reflected light vector are calculated by solving the binary simultaneous equations, so that the processing for creating the above-described reflection model can be speeded up. Further, each vector obtained by the above-described vector determination means 112 and the vector generation means 111 and the coefficient obtained by the coefficient calculation means 113 are stored in the vector holding means 10.
3 and the coefficient holding means 104 respectively. Therefore, when these vectors and coefficients are changed by the changing unit 106, the image data is newly calculated by the image data calculating unit 105 in accordance with the above-mentioned reflection model, sent to the image storing unit 101, and the object included in the image is calculated. Color and surface properties can be changed. In this manner, a reflection model is created using the method of claim 1, and based on this reflection model, it is possible to realize an image data manipulation device that changes the color and surface properties of an object. Such change processing can be speeded up.

[0022]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 3 shows the configuration of an embodiment of the image data manipulation device of the present invention.

In FIG. 3, image data input from an image input device 201 such as an image scanner is held in a holding memory 202 corresponding to the image storing means 101.
The display device 205 performs display based on the image data transferred from Step 2 to the display memory 204.

The image data operation device is provided with an input device 206 such as a mouse or a light pen. When an operator operates the input device 206 to specify a pixel on the screen, the input device control is performed. The unit 207 is configured to send the coordinates of the corresponding pixel to the readout circuit 211. Also, icons or the like indicating various instructions are displayed on the screen of the display device 205, and when the operator designates these icons via the input device 206, the input device control unit 207 displays the corresponding instructions. What is necessary is just to make it the structure which recognizes an input.

The object extracting unit 208 extracts the image data of each pixel included in the object portion from the image held in the holding memory 202 and transfers the address of the holding memory 202 corresponding to the pixel. The signal is sent to the circuit 203. In response, the transfer circuit 203
By transferring the corresponding image data from the holding memory 202 to the display memory 203, the display device 2
At 05, the image of the part of the object to be processed is displayed.

The reading circuit 211 reads out the image data of the pixel indicated by the input coordinates from the holding memory 202 and sends it to the vector determining unit 212.

Here, the above-mentioned Japanese Patent Application No. 3-195326 is disclosed.
In order to use the signal technique, the object color vector, the illumination light vector, and the secondary reflected light vector must be determined in advance.

As the object color vector, a representative color of the target object may be designated. So, for example,
The operator only has to subjectively determine the pixel indicating the color of the object itself and specify the corresponding pixel. In response, the corresponding coordinates are sent from the input device control unit 207 to the readout circuit 211, and the readout circuit 211 reads out the image data of the corresponding pixel and sends it to the vector determination unit 212. The vector determining unit 212 may transmit the image data obtained in this way as it is to the vector holding memory 214 corresponding to the vector holding unit 103 as an object color vector.

By the way, in both natural images such as photographs and computer graphics, sunlight, fluorescent light or light from a white light bulb is often used as illumination light. That is, light generally used as illumination is light close to white light, and the values of the RGB components of the illumination light vector are close to each other. Therefore, assuming that the illumination light is ideal white light, it is possible to approximate all the components of the illumination light vector with the same value.

Further, since the secondary reflected light depends on the color of the illumination light, similarly, the secondary reflected light is regarded as an ideal white light, and the values of the respective components of the secondary reflected light vector are all the same. It can be approximated by value. In addition, since the magnitude of the secondary reflected light vector is small, the contribution to each component of the image data is also small. Therefore, when the illumination light is other than white light, even if the secondary reflected light is approximated by ideal white light, the image quality does not significantly deteriorate.

From the above, it can be seen that at least one of the illumination light vector and the secondary reflected light vector can be approximated by ideal white light. For example, the memory 213
In advance, the illumination light vector and the secondary reflection light vector when the illumination light and the secondary reflection light are ideal white light are held in advance, and the vector determination unit 212 responds to the instruction from the operator. One of them may be read from the memory 213 and input to the vector holding memory 214 described above.

In this case, the value of the RGB component indicating high luminance is stored in the memory 213 as the illumination light vector, and the value of the RGB component indicating relatively low luminance is stored as the secondary reflected light vector. Is the input device 20 described above.
6, it may be indicated that the illumination light or the secondary reflected light is approximated by ideal white light. Further, in addition to the above-described approximate illumination light vector and secondary reflected light vector,
Memory 2 stores illumination light vectors and secondary reflected light vectors having component ratios corresponding to typical light sources such as sunlight and white light.
13 and the vector determination unit 212 may select the corresponding vector as the vector that is not approximated.

As described above, the readout circuit 211, the vector determination section 212, and the memory 213 operate in response to an instruction from the input device 206, whereby the vector determination means 1
12 and the function of the vector generation means 111,
An object color vector, an illumination light vector, and a secondary reflected light vector can be obtained.

Next, a method for creating a reflection model of a target object using the object color vector, the illumination light vector, and the secondary reflected light vector obtained in this manner will be described.

Here, in the above equation, for example, on the condition that each component of the illumination light vector (R _s , G _s , B _s ) is equal (that is, R _s = G _s = B _s ), Clearing the _{terms, R 0 -G 0 = K 1} (R d -G d) + K 3 (R a -G a) ... G 0 -B 0 = K 1 (G d -B d) + K 3 (G a −B _a )... By applying the Cramer's formula to the system of equations consisting of these equations, the coefficients K ₁ and K ₃ are

[0036]

(Equation 02)

[0037]

[Equation 03]

Are represented as follows. By substituting the coefficients K ₁ and K ₃ obtained in this way into the equation,
It is possible to obtain the coefficients K ₂ corresponding to the illumination light. In this case, the coefficient calculating unit 113 substitutes the image data, the object color vector, and the secondary reflected light vector into the above-described expression according to the input of the image data of each pixel of the object portion. calculating the coefficients K _1, K _3, it may be configured to calculate the coefficient K ₂ based on the obtained coefficient K _1, K _3.

Similarly, the secondary reflected light vector (R _a ,
G _a , B _a ), if the components are equal, the equation is transformed, and the Camel equation is applied to the obtained system of binary equations, the coefficients K ₁ and K _{2 become}

[0040]

[Equation 04]

[0041]

[Equation 05]

The obtained coefficient K is expressed as
Based on ₁ and the coefficient K _2, it can be obtained coefficient K _3, corresponding to the secondary reflected light. In this case, the coefficient calculating unit 113 substitutes the image data, the object color vector and the illumination light vector into the above-described expression according to the input of the image data of each pixel of the object portion, and obtains the coefficient K
_1, K ₂ is calculated, and may be configured to calculate the coefficient K ₃ on the basis of the obtained coefficient K _1, K _2.

For example, as shown in FIG. 3, the coefficient calculating means 113 is composed of a microprocessor (MPU) 221 and a ROM.
In the case where the illumination light is approximated by white light and the case where the secondary reflected light is approximated by white light, the calculation procedure is respectively made into firmware, and the calculation procedure is performed according to an instruction from the operator. The configuration may be such that the calculation procedure is selected and executed. Further, the image data extracted by the object extracting unit 208 may be sequentially input to the microprocessor 222.

Here, for example, when the illumination light is approximated by white light, the number of times of the multiplication processing and the division processing required for obtaining the coefficient K ₁ is five, and the ternary simultaneous processing is performed without using the approximation processing. It can be seen that the number is significantly reduced as compared with 25 times, which is the number of times of the multiplication / division processing required when solving the equation.

Therefore, as described above, the processing for calculating the coefficients K ₁ , K ₂ , and K ₃ is simplified on the condition that the illumination light or the secondary reflected light is approximated by ideal white light. , K ₁ , K ₂ , K corresponding to each pixel
₃ Significantly reduced the amount of computation required to calculate ₃ ,
The time required for the coefficient calculation process can be reduced.

The coefficients K ₁ and K obtained as described above
₂ and K ₃ are sequentially sent out and held in a coefficient holding memory 215 corresponding to the coefficient holding means 104. The coefficient holding memory 215 stores, for example, the object extracting unit 208 described above.
Receives the notification of the address of the holding memory 202 corresponding to each extracted pixel, and inputs the input coefficients K ₁ , K ₂ , K
₃ may be stored in correspondence with the notified address.

In this way, the reflection model representing the color of each pixel of the object portion in the image can be created by using the method of the first aspect. Therefore, by substituting the coefficients K ₁ , K ₂ , and K ₃ corresponding to the respective pixels held in the above-described coefficient holding memory 215 and the respective vectors held in the vector holding memory 214 into the expression, each of the parts of the object can be obtained. The image data of the pixel is reproduced.

For example, the arithmetic circuit 231 and the read circuit 232
Form image data calculation means 105, and readout circuit 23
2, the coefficients K ₁ , K ₂ , and K ₃ corresponding to each pixel are read from the above-described coefficient holding memory 214, and the arithmetic circuit 2
In response, the arithmetic circuit 231 calculates the image data of each pixel according to the equation, and outputs the image data to the display memory 20.
4 may be sent.

Further, by changing the value of each vector held in the vector holding memory 214 in accordance with the instruction of the operator, the color of the object and the color of the illumination can be changed. For example, when changing the color of an object, the display device 20
5, a color palette is displayed, and when the operator designates a pixel in the area where the color palette is displayed via the input device 206, the input device control unit 207 rewrites the coordinates of the corresponding pixel. 241 may be specified to specify the changed target color. The rewrite processing unit 241 reads the corresponding image data from the display memory 204 based on the notified coordinates, and rewrites the value of the object color vector held in the vector holding memory 214 using the image data. .

As described above, according to the instruction from the operator,
When the input device control unit 207 and the rewriting processing unit 241 operate to change the object color vector, the changing unit 11
6 can be realized, and the color of the object can be freely changed while maintaining the texture of the object. At this time, since the color of the object displayed by the display device 205 is determined according to the reflection model shown in the equation, it is possible to obtain a high-quality image that does not make human eyes feel unnatural. As a result, it is possible to easily realize a simulation in a case where the color of the paint or the color of the pigment mixed with the plastic material is variously changed.

Similarly, the illumination light vector and the secondary reflected light vector can be changed, and a high-quality image can be obtained without causing unnaturalness to human eyes. In this case, a simulation of a state where an object is illuminated by light sources having various spectra can be easily realized.

Further, in accordance with the instruction of the operator, the coefficients K ₁ and K ₁ corresponding to each pixel held in the coefficient holding memory 215.
By changing the value of _2, K _3, it is possible to vary the properties and material feeling of the surface of the object.

In general, the diffuse reflection characteristic represented by the coefficient K ₁ monotonously decreases as the angle θ between the normal line on the surface of the object and the direction of the line of sight increases. But,
For example, by changing the value of the coefficient K ₁ representing the diffuse reflection characteristics,
If the diffuse reflection characteristic of the object has directivity, it is possible to obtain a shiny material feeling like mica.

In this case, in response to an instruction from the operator, the characteristic setting section 242 sets a diffuse reflection characteristic such that the diffuse reflection intensity has an extreme value at a specific angle θ. What is necessary is just to adopt a configuration in which a difference from the reflection characteristic is obtained for each pixel, and the rewrite processing unit 241 is notified. In response to this, the rewrite processing unit 241 rewrites the value of the coefficient K ₁ in the coefficient holding memory 215 to change the material appearance of the object to a material appearance having diffuse reflection characteristics with directivity. it can.

That is, the function of the changing means 116 is performed by the rewriting processing section 241 and the characteristic setting section 242. Similarly, it is possible to change the coefficient K ₁ depending on the diffuse reflection characteristics corresponding to the various material appearance, change the material appearance of an object while keeping the properties of the object color and the surface,
It is possible to obtain a high-quality image that does not make human eyes feel unnatural.

Similarly, by changing the coefficient K ₂ representing the specular reflection characteristic, the glossiness of the surface of the object can be changed, and the surface roughness of the object can be changed while preserving the color and material of the object. Can be changed.

Further, by changing the coefficient K ₃ representing the secondary reflection characteristic, it is possible to obtain the effect of changing the distance from the surrounding object by changing the intensity of the secondary reflected light component. In each case, these coefficients K ₁ ,
Even when K ₂ and K ₃ are changed in combination, it is possible to obtain a high-quality image that does not make human eyes feel unnatural.

That is, as described above, a reflection model is created by approximating the illumination light or the secondary reflected light with white light, and the color, material appearance, surface properties, etc. of the object are obtained using this reflection model. It is possible to speed up the process of freely changing and with high accuracy.

As a result, it is possible to reduce the user's waiting time when changing the color and the surface properties of the object in the image, and to improve the efficiency of the work of designing products and posters and the work of creating computer graphics. It is possible to provide an image processing system that is easy for the user to use.

[0060]

As described above, the first aspect of the present invention is:
By approximating the illumination light or the secondary reflection light with ideal white light, the process of calculating the coefficients to be multiplied by the object color vector, the illumination light vector, and the secondary reflection light vector is simplified, and the reflection model of the object can be made faster. Can be created. Therefore, it is possible to speed up the entire process of freely and highly accurately changing the color, texture, and surface properties of an object using this reflection model.

According to a second aspect of the present invention, a reflection model is created by the reflection model creating method of the first aspect, and based on the reflection model, the color, texture, and surface properties of the object can be freely and highly improved. It is possible to speed up the process of changing with accuracy. As a result, it is possible to reduce the user's waiting time when changing the color and the surface properties of the object in the image, and to improve the efficiency of the design work of products and posters and the creation work of computer graphics.

[Brief description of the drawings]

FIG. 1 is a diagram showing the principle of a reflection model creating method according to claim 1;

FIG. 2 is a diagram showing a configuration of an image data manipulation device according to a second embodiment.

FIG. 3 is a block diagram of an embodiment of an image data manipulation device according to claim 2;

[Explanation of symbols]

 Reference Signs List 101 Image storage means 102 Reflection model creation means 103 Vector holding means 104 Coefficient holding means 105 Image data calculation means 106 Changing means 111 Vector generation means 112 Vector determination means 113 Coefficient calculation means 201 Image input device 202 Holding memory 203 Transfer circuit 204 Display Memory for display 205 Display device 206 Input device 207 Input device control unit 208 Object extraction unit 211, 232 Readout circuit 212 Vector determination unit 213 Memory 214 Vector storage memory 215 Coefficient storage memory 221 Microprocessor (MPU) 222 ROM 231 Operation circuit 241 Rewriting process Unit 242 Characteristic setting unit

Claims (2)

(57) [Claims] 1. An image processing apparatus according to claim 1, wherein each component obtained by multiplying a vector representing the color of the object itself, the color of the illumination light, and the color of the secondary reflected light included in the image by a coefficient corresponding to each vector. In a reflection model creating method for obtaining a reflection model representing image data representing a color of each pixel of a portion, an illumination light vector indicating an illumination light color and a secondary reflection light indicating a color of a secondary reflected light based on the image data. One of the secondary reflected light vector and the object color vector indicating the color of the object itself are determined, and one of the illumination light vector or the secondary reflected light vector is determined as a vector corresponding to an ideal white light. Approximately using the expression representing the reflection model to eliminate a component including a vector corresponding to the ideal white light, and to remove the object color vector and the illumination light vector or Derives a binary simultaneous equation with the secondary reflected light vector, solves the binary simultaneous equation, and calculates the object color vector, the illumination light vector, and the coefficient corresponding to the secondary reflected light vector, respectively. A method for creating a reflection model, characterized in that it is obtained. 2. An image storage means (101) for storing image data representing a color of each pixel constituting an image, and an object color vector indicating a color of an object itself included in the image based on the image data. Reflection model creation means (102) for obtaining an illumination light vector indicating the color of the illumination light, a secondary reflection light vector indicating the color of the secondary reflection light, and a coefficient corresponding to each of these vectors; A vector holding unit (103) for holding an illumination light vector and the secondary reflected light vector, a coefficient holding unit (104) for holding a coefficient corresponding to each vector for each pixel, and the object color vector And the illumination light vector and the secondary reflected light vector are multiplied by corresponding coefficients to calculate image data of each pixel. Image data calculating means (105) to be transmitted to the CPU, and in response to the input of the change instruction, the values of the object color vector, the illumination light vector, the secondary reflected light vector, and the coefficients respectively corresponding to these vectors are calculated. In the image data manipulation device provided with changing means (106) for changing, the reflection model creating means (102) receives an approximation instruction for designating one of the illumination light vector and the secondary reflected light vector. A vector generating means (111) for generating a vector indicating an ideal white light as a corresponding vector and transmitting the vector to the vector holding means (103), and the illumination light vector or the The secondary reflected light vector and the object color vector are determined based on image data stored in the image storage means (101). A vector determining means (112) to be transmitted to the vector holding means (103); and a predetermined binary simultaneous equation relating to the illumination light vector or the secondary reflected light vector and the object color vector according to the approximation instruction. The vector determination means (11
Substituting each vector and the image data obtained in 2), obtaining a coefficient corresponding to these vectors and a coefficient corresponding to the vector specified by the approximation instruction, and sending them to the coefficient holding means (104). Coefficient calculating means (11
3) An image data operation device characterized by comprising: