JP2022065752A - Image processing apparatus, image processing method, and program - Google Patents

Abstract

To provide processing for obtaining the distribution characteristics of bright material included in an object.SOLUTION: An image processing apparatus has: input means that receives input of image data obtained by picking up an image of an object including bright material; acquisition means that acquires condition information representing a geometric condition in picking up the image of the object; extraction means that, based on the image data, extracts color information on the bright material in a target area of the object; and derivation means that, based on the condition information and the color information on the bright material, derives the distribution characteristics of the bright material in the target area.SELECTED DRAWING: Figure 8

Description

The present invention relates to a technique for deriving characteristics relating to a bright material contained in an object.

When metallic coating or pearl coating is applied to industrial products, a paint containing a bright material is used. Patent Document 1 is a technique for evaluating the surface of an object painted with a paint containing a bright material. Patent Document 1 discloses a technique for separating and evaluating the color of a bright material and the color of a base material based on an image of reflected light on a painted surface.

Japanese Unexamined Patent Publication No. 2019-78623

When evaluating a painted surface, it may be desired to know not only the color of the bright material but also the distribution characteristics of the bright material, but Patent Document 1 has a problem that the distribution characteristics of the bright material cannot be obtained. ..

Therefore, an object of the present invention is to provide a process for obtaining distribution characteristics of a bright material contained in an object.

In order to solve the above problems, the image processing apparatus according to the present invention has an input means for inputting image data obtained by imaging an object containing a bright material, and condition information representing geometric conditions when imaging the object. Based on the acquisition means for acquiring the image data, the extraction means for extracting the color information of the bright material in the target region of the object, and the condition information and the color information of the bright material, the target. It is characterized by having a derivation means for deriving the distribution characteristics of the bright material in the region.

It is possible to obtain the distribution characteristics of the bright material contained in the object.

Diagram to explain the principle of thin film interference The figure which shows the hardware configuration of an image processing apparatus. The figure which shows the functional composition of an image processing apparatus. Diagram showing an example of a user interface Diagram showing an example of a histogram Diagram showing geometric conditions when imaging an object The figure which shows the rotation angle of a bright material A flowchart showing the processing executed by the image processing device. Diagram showing an example of a color chart The figure for demonstrating the optical path of the light reflected by the bright material. Diagram showing an example of a user interface Diagram showing geometric conditions when imaging an object The figure which shows the functional composition of an image processing apparatus. A flowchart showing the processing executed by the image processing device. The figure for demonstrating the distribution characteristic of a bright material on a painted surface.

Hereinafter, each embodiment will be described with reference to the drawings. The following embodiments do not necessarily limit the present invention. Moreover, not all combinations of features described in each embodiment are essential for the means of solving the present invention.

[First Embodiment]
Some of the bright materials contained in the paint used for painting industrial products change their colors depending on the observation conditions such as the observation angle and the irradiation angle of light. Typical bright materials include interference mica in which mica is coated with titanium oxide or the like, and silica flakes in which silica is coated with titanium oxide or the like. The color of such a bright material changes depending on the observation conditions due to thin film interference. It is known that the brilliant material that utilizes thin film interference changes its brilliance according to the distribution in the paint. FIG. 15 is a diagram showing two types of coating in a simple manner, and the coated surface contains a bright material. When the orientation of the bright material is uniform as in the coating shown in FIG. 15 (a), the colors are the same among the bright materials existing in the vicinity. On the other hand, when the orientation of the bright material is non-uniform as in the coating shown in FIG. 15 (b), the light incident angle and the observation angle with respect to each bright material are different, so that the bright material has various colors. appear. Further, since the number of bright materials differs between FIGS. 15 (a) and 15 (b), a characteristic called a bright feeling, in which the position of the bright spot changes brilliantly, is also different. That is, even if they are painted with the same paint, they will look different if the orientation and number of bright materials differ depending on the position.

Therefore, in the present embodiment, information on the distribution characteristics of the bright material on the painted surface, such as the orientation and the number of the bright materials, is derived based on the image data obtained by imaging the object. The target object in the present embodiment is an object painted with a paint containing a bright material, and a bright material whose color changes according to observation conditions due to thin film interference is used.

First, the principle of thin film interference that occurs in a bright material will be described with reference to FIG. As shown in FIG. 1, a bright material utilizing thin film interference is made by coating a substrate such as mica or silica with a medium having a high refractive index. When light is incident on the bright material, a part of the incident light is reflected on the upper surface of the thin film. Further, a part of the incident light is incident on the thin film, refracted, and then reflected on the base material. Then, outside the thin film, the light is received by an image pickup device or a person in a state where these two reflected lights interfere with each other. In this case, assuming that the thickness of the coating layer is d, the refractive index is n, and the refraction angle is γ, the optical path difference between the two reflected lights is 2dncosγ. A color corresponding to the wavelength λ at which this optical path difference is (m + 1/2) λ is observed by an image pickup apparatus or a human being. That is, a color corresponding to the wavelength λ satisfying the following equation (1) is observed.

Here, m = 0, 1, 2, .... For example, in the case of interference mica, the mica is coated with titanium dioxide having a refractive index of about 2.5 so that the thickness of the coating layer is about 80 nm. Only when m = 0, λ is the wavelength of visible light. In this embodiment, a bright material controlled so that λ is the wavelength of visible light is used only when m = 0. Substituting m = 0 into equation (1), λ is expressed as in equation (2).

<Hardware configuration of image processing device>
The hardware configuration of the image processing apparatus 1 in the present embodiment will be described with reference to FIG. In FIG. 2, the image processing device 1 includes a CPU 201, a ROM 202, and a RAM 203. Further, the image processing device 1 includes a VC (video card) 204, a general-purpose I / F (interface) 205, a SATA (serial ATA) I / F 206, and a NIC (network interface card) 207.

The CPU 201 uses the RAM 203 as a work memory to execute an OS (operating system) and various programs stored in the ROM 202, the HDD (hard disk drive) 212, and the like. Further, the CPU 201 controls each configuration via the system bus 208. The process according to the flowchart described later is executed by the CPU 201 after the program code stored in the ROM 202, the HDD 212, or the like is expanded in the RAM 203. A display device 2 is connected to the VC204. An input device 210 such as a mouse or a keyboard and an image pickup device 3 are connected to the general-purpose I / F 205 via a serial bus 209. A general-purpose drive 213 for reading and writing HDD 212 and various recording media is connected to SATAI / F206 via a serial bus 211. The NIC 207 inputs and outputs information to and from an external device. The CPU 201 uses various recording media mounted on the HDD 212 and the general-purpose drive 213 as storage locations for various data. The CPU 201 displays a GUI (graphical user interface) provided by the program on the display device 2, and receives an input such as a user instruction received via the input device 210.

<Functional configuration of image processing device>
FIG. 3 is a block diagram showing a functional configuration of the image processing device 1. The CPU 201 functions as the functional configuration shown in FIG. 3 by reading and executing a program stored in the ROM 202 or the HDD 212 using the RAM 203 as a work memory. It should be noted that it is not necessary that all of the processes shown below are executed by the CPU 201, and even if the image processing device 1 is configured so that a part or all of the processes are performed by one or a plurality of processing circuits other than the CPU 201. good.

The image processing device 1 has an input unit 301, a setting unit 302, an extraction unit 303, an acquisition unit 304, a derivation unit 305, and an output unit 306. The input unit 301 inputs image data obtained by imaging an object object including a bright material. The image represented by the image data in the present embodiment has RGB values (R, G, B) as pixel values for each pixel. The setting unit 302 sets the target area in the image represented by the input image data based on the instruction of the user. The extraction unit 303 extracts the color information of the bright material from the target area set by the setting unit 302. The color information of the bright material in the present embodiment is an XYZ value histogram of the bright spots included in the image. An example of the XYZ value histogram is shown in FIG. In the graph of FIG. 5, each of the X value, the Y value, and the Z value is the horizontal axis, and the vertical axis is the frequency.

The acquisition unit 304 acquires condition information representing geometric conditions when imaging the target object. The acquisition unit 304 in the present embodiment acquires the light receiving angle α as condition information. The light receiving angle α will be described with reference to FIG. FIG. 6 shows the geometric conditions when an object is imaged by using the image pickup apparatus 3. In FIG. 6, the light reflected by the bright material contained in the target object is incident on the image pickup apparatus 3 as shown by an arrow. When a perpendicular line is drawn with respect to the painted surface of the target object at the position where the light is reflected, the angle formed by the perpendicular line and the reflected light is the light receiving angle α. In the present embodiment, it is assumed that the light receiving angle α is uniform within the target region set by the setting unit 302. Specifically, the angle formed by the perpendicular line and the reflected light with respect to the painted surface at the center of the target area is defined as the light receiving angle α. It is not always necessary to use the light receiving angle α at the center of the target region, and for example, the light receiving angle α at a point other than the center in the target region may be used.

The derivation unit 305 derives the distribution characteristics of the bright material by fitting the XYZ value histogram calculated using the parameters of the model relating to the distribution characteristics of the bright material to the XYZ value histogram extracted by the extraction unit 303. .. As described with reference to FIG. 15, since the orientation and the number of the bright materials affect the appearance of the object, the orientation and the number of the bright materials are modeled by a plurality of parameters. The number of bright materials can be modeled with one parameter, M. A model of the orientation of the bright material will be described with reference to FIG. As shown in FIG. 7, the bright material is rotated clockwise by θ degrees with respect to the direction parallel to the painted surface. The orientation of the bright material is modeled by assuming the distribution of the rotation angle θ of each bright material. For example, it is assumed that the angle of rotation θ of the bright material follows a probability density function of normal distribution as shown in Eq. (3). In this case, the mean μ and the standard deviation σ are the parameters for the model of the orientation of the bright material. In the equation (3), the probability density function of the angle of rotation θ is f (θ).

A method of calculating the XYZ value histogram using the above-mentioned parameters (μ, σ, M) will be described with reference to FIG. FIG. 10 shows the geometric conditions when the incident light is reflected by the bright material and the reflected light is received by the image pickup apparatus 3. Let β be the angle formed by the incident light and the perpendicular to the surface of the covering layer. Since a medium having a high refractive index is used for the coating layer, the incident light entering the coating layer is refracted. Let this refraction angle be γ. The refracted light is reflected at the substrate. The angle formed by the reflected light and the perpendicular line to the painted surface is α', and the angle formed by the reflected light refracted when emitted from the painted surface and the perpendicular line to the painted surface is the above-mentioned light receiving angle α. Assuming that the refractive index of the paint contained in the painted surface is n and the refractive index of air is 1, the equation (4) holds according to Snell's law.

Similarly, if the refractive index of the coating layer is n2, the equation (5) holds.

In addition, equation (6) also holds.

Substituting Eq. (6) into Eq. (5), Eq. (7) holds.

Substituting Eq. (7) into Eq. (2), Eq. (8) holds.

Based on the equations (4) and (8), the refractive index of the paint is observed by the image pickup apparatus 3 using n, the thickness d of the coating layer, the light receiving angle α, and the rotation angle θ. The wavelength λ corresponding to the color can be calculated. In the present embodiment, the refractive indexes n and n2 and the thickness d of the coating layer are predetermined constants. The derivation unit 305 calculates the XYZ value histogram by applying the color matching function to the calculated wavelength λ, and adjusts the parameters (μ, σ, M), and the XYZ value histogram extracted by the extraction unit 303. Make a fitting for.

Specifically, the derivation unit 305 has an arbitrary number N pieces according to the probability density function f (θ) in the range of (−π, π) for a certain parameter combination (μ', σ', M'). Generates the random number of θ as θ. Further, by substituting θ into Eq. (8), N λs can be calculated. When calculating λ, α'is calculated using the equation (4) based on the light receiving angle α acquired by the acquisition unit 304, and is substituted into the equation (8). The derivation unit 305 calculates the XYZ value histogram by applying the color matching function to the calculated wavelength λ. However, since this XYZ value histogram has an area corresponding to an arbitrary number of N pieces, the XYZ value histogram is multiplied by M / N in order to have an area corresponding to M parameters.

The derivation unit 305 adjusts the parameters (μ, σ, M) so that the calculated XYZ value histogram has a high degree of similarity to the XYZ value histogram extracted by the extraction unit 303. The derivation unit 305 calculates the area of the region where the histograms overlap and the Kullback-Leibler divergence for each of the histograms of the X value, the Y value, and the Z value, and sets the total value as the similarity. The derivation unit 305 calculates the similarity for a combination of several parameters, and uses the combination of the parameters corresponding to the highest similarity among the calculated combinations of the parameters as the derivation result of the distribution characteristic of the bright material. .. The combination of parameters may be determined by using a known optimization method such as a genetic algorithm method, a Levenberg-Marquardt method, or a Gauss-Newton method.

The output unit 306 outputs distribution information representing the distribution characteristics of the bright material derived by the derivation unit 305.

<Processes executed by the image processing device>
FIG. 8 is a flowchart showing a process executed by the image processing device 1. Hereinafter, each step (process) is represented by adding S in front of the reference numeral. In S801, the input unit 301 inputs image data based on the user's instruction. FIG. 4 shows an example of a UI for accepting a user's instruction. The input area 401 is an area for the user to input a file path to the image data stored in the HDD 212 or the like. The input unit 301 reads the image data indicated by the input file path, and displays the image represented by the image data in the image display area 402.

In S802, the setting unit 302 sets the target area in the image represented by the input image data based on the instruction of the user. Specifically, the setting unit 302 receives the user's area designation using the input device 210 in the image display area 402. The method of setting the target area is not limited to the method based on the user's instruction. For example, the position information of the target area in the image may be saved in the HDD 212 or the like in advance, and the setting unit 302 may read the position information from the HDD 212 or the like. .. According to this method, in a factory line inspection or the like in which a desired target area is installed at the same position every time, the target area can be set without the user having to specify the area many times. The target area is set in the shape of a circle, a rectangle, or the like.

In S803, the extraction unit 303 converts the pixel values of the pixels included in the target area set in S802. Specifically, the extraction unit 303 converts RGB values (R, G, B) into XYZ values (X, Y, Z) using the following equation (9).

The coefficients a0 to a8 included in the matrix used as information indicating the color conversion rule are derived in advance by, for example, imaging with the image pickup apparatus 3 and measurement with a spectral radiance meter for the color chart shown in FIG. Keep it. The imaging and measurement of the color chart are performed in the same lighting environment as the lighting environment when the target object is imaged. When deriving the coefficients a0 to a8, the average value of the RGB values is calculated for each of the 24 patches in the captured image. Coefficients a0 to a8 are derived by the least squares method so that the XYZ value obtained by converting each of the calculated average values of the RGB values by the equation (9) approaches the measured value (XYZ value) of the corresponding patch.

In S804, the extraction unit 303 detects a bright spot in the target region. Specifically, the extraction unit 303 sets the average value of the Y values in the target region as a threshold value, and performs binarization processing on the Y value of each pixel using the set threshold value. By the binarization process, the Y value becomes 1 indicating a high-luminance region or 0 indicating a non-luminance region. The extraction unit 303 detects a pixel having a Y value of 1 as a pixel constituting a bright spot. When the bright spot is composed of a plurality of pixels, the connected component is calculated, and the average value of each of the X value, the Y value, and the Z value in the connected component is calculated. In this embodiment, since the Y value is binarized only for detecting the bright spot, the Y value used for the calculation of the average value in the connected components and the subsequent processing is before binarization. The value is used.

In S805, the extraction unit 303 extracts the color information of the bright material based on the bright spots detected in S804. Specifically, the extraction unit 303 creates a histogram of each of the X value, Y value, and Z value of the bright spot. Here, the area of each of the XYZ value histograms is the number of detected bright spots.

In S806, the acquisition unit 304 acquires the condition information representing the geometrical condition when the target object is imaged. Specifically, the acquisition unit 304 acquires the light receiving angle α determined by the target region set in S802 and the orientation and positional relationship of the image pickup apparatus 3. The acquisition unit 304 in the present embodiment acquires the value input by the user as the light receiving angle α. As in the setting of the target area, the light receiving angle α measured by using the measuring instrument may be stored in the HDD 212 or the like in advance, and the acquisition unit 304 may read the light receiving angle α from the HDD 212 or the like. Further, when the target object faces the image pickup apparatus 3, the acquisition unit 304 may calculate the light receiving angle α based on the image. Specifically, the acquisition unit 304 acquires the distance [mm] on the sensor by calculating the distance [Pixel] between the center of the target area and the principal point in the image and multiplying by the pixel pitch. The acquisition unit 304 calculates arctan (distance [mm] on the sensor / focal length [mm]) as the light receiving angle α. Further, for example, a three-sided survey may be performed using a Time Of Flight (TOF) sensor or the like to calculate the light receiving angle α.

In S807, the derivation unit 305 fits the XYZ value histogram calculated by using the parameters of the model regarding the distribution characteristics of the bright material to the XYZ value histogram extracted as the color information in S805. The derivation unit 305 derives the parameters (μ, σ, M) as distribution information representing the distribution characteristics of the bright material by fitting the histogram. In S808, the output unit 306 outputs distribution information representing the distribution characteristics of the bright material derived by the extraction unit 305 to the display device 2. Specifically, the output unit 306 displays the parameters (μ, σ, M) in the result display area 403 shown in FIG. The method of outputting the distribution information is not limited to the method of displaying on the display device 2, and may be output by writing to, for example, a csv file or a txt file stored in the HDD 212 or the like.

<Effect of the first embodiment>
As described above, the image processing apparatus in the present embodiment inputs image data obtained by imaging an object containing a bright material. Acquires condition information that represents geometric conditions when imaging an object. Based on the image data, the color information of the bright material in the target area of the object is extracted. Based on the condition information and the color information of the bright material, the distribution characteristics of the bright material in the target region are derived. Thereby, the distribution characteristic of the bright material contained in the object can be obtained.

<Modification example>
In the present embodiment, the two parameters for specifying the distribution characteristics of the bright material are the orientation of the bright material and the number of the bright materials, but only the orientation of the bright material may be used. That is, the fitting parameter is (μ, σ). In this case, since it is necessary to align the areas of the histograms before calculating the similarity, each histogram is normalized so that the area becomes 1.

Further, in the present embodiment, a bright material controlled so that λ is the wavelength of visible light is used only when m = 0, but λ is the wavelength of visible light even when m = 1 or more. A bright material controlled in such a manner may be used. In this case, m = 0, 1, 2, ... Is substituted into the equation (1) to calculate the wavelength λ included in the visible light range.

Further, the extraction unit 303 in the present embodiment calculates the average value of each of the X value, the Y value, and the Z value in the connected component when the bright spot is composed of a plurality of pixels, but the statistic is an average value. It may be the median value or the like. In this case, it is difficult to normalize the histogram by the number of bright materials. Therefore, the similarity is calculated after normalizing each histogram so that the area becomes 1.

Further, although the derivation unit 305 in the present embodiment uses a model of the distribution of the angle of rotation θ as a normal distribution, other distributions may be used. For example, if it is assumed that a plurality of types of bright materials are contained in the paint, a mixed Gaussian distribution may be used, or a von Mises distribution may be used.

Further, when the derivation unit 305 uses an optimization method such as the Levenberg-Marquardt method, the solution may not be determined if there is only one target area for a plurality of parameters. In this case, the target area set by the setting unit 302 may be divided, and the number of data may be secured by calculating the XYZ value histogram of each area obtained by the division. Further, the setting unit 302 may set a plurality of target areas based on the user's instruction.

Further, in the present embodiment, the XYZ value is used as the color information, but a value in another color space such as an RGB value or an L ^* a ^* b ^* value may be used.

Further, in the present embodiment, the image represented by the input image data has an RGB value for each pixel, but a value in another color space may be used as a pixel value. Further, it may have a value corresponding to a wavelength region different from RGB such as spectral information or a value of 3 channels or more as a pixel value.

[Second Embodiment]
In the first embodiment, the distribution characteristics of the bright material contained in the target region were derived and output to the display device. In this embodiment, the difference in the distribution characteristics of the bright material is evaluated between two target regions having different observation conditions. Since the hardware configuration of the image processing apparatus in this embodiment is the same as that in the first embodiment, the description thereof will be omitted. Hereinafter, the differences between the present embodiment and the first embodiment will be mainly described. The same configuration as that of the first embodiment will be described with the same reference numerals.

<Evaluation of distribution characteristics of bright materials>
As shown in FIG. 11, the difference in the distribution characteristics of the bright material between the two regions is evaluated by comparing the color information of the bright material between the region 1101 in the target object and the region 1102 different from the region 1101. Think about it. FIG. 12 shows the geometrical conditions for imaging each target area using the image pickup apparatus 3. As shown in FIG. 12, the light receiving angle α1 corresponding to the region 1101 and the light receiving angle α2 corresponding to the region 1102 are different. As shown in the formulas (4) and (8), the observed color of the bright material is affected by the light receiving angle α and the rotation angle θ. When comparing the color information of the bright material between the two regions, it is not possible to specify whether the difference is due to the difference in the light receiving angle or the difference in the distribution characteristics of the bright material.

Therefore, in the present embodiment, the distribution characteristics of the bright material in the region 1101 are derived, and based on the derived distribution characteristics, the bright material in the case where the region 1101 is imaged under the observation conditions (light receiving angle α2) corresponding to the region 1102. Color information (hereinafter referred to as virtual color information) is calculated. By comparing the calculated virtual color information of the bright material in the region 1101 and the color information of the bright material in the region 1102, the difference in the distribution characteristics of the bright material between the two regions is evaluated. In this way, by matching the observation conditions and comparing the color information of the bright materials, it is possible to evaluate the difference in the distribution characteristics of the bright materials between the two regions with high accuracy.

<Functional configuration of image processing device>
FIG. 13 is a block diagram showing a functional configuration of the image processing device 1. The image processing device 1 has an input unit 301, a setting unit 302, an extraction unit 303, an acquisition unit 304, a derivation unit 305, a calculation unit 1301, an evaluation unit 1302, and an output unit 306. The calculation unit 1301 calculates the virtual color information of the bright material in the region 1101 based on the distribution characteristics of the bright material in the region 1101 and the geometric conditions corresponding to the region 1102. The evaluation unit 1302 calculates the similarity between the virtual color information of the bright material in the region 1101 and the color information of the bright material in the region 1102 as an evaluation value.

<Processes executed by the image processing device>
FIG. 14 is a flowchart showing a process executed by the image processing device 1. In S1401, the input unit 301 inputs image data based on the user's instruction. In S1402, the setting unit 302 sets the area 1101 and the area 1102 as the target area in the image represented by the input image data based on the instruction of the user.

In S1403, the extraction unit 303 converts RGB values (R, G, B) into XYZ values (X, Y, Z) for each of the target areas set in S1402. In S1404, the extraction unit 303 detects bright spots in each of the target regions. In S1405, the extraction unit 303 extracts the color information of the bright material corresponding to each target region based on the bright spots detected in S1404.

In S1406, the acquisition unit 304 acquires the condition information representing the geometrical condition when the target object is imaged for each target area. Specifically, the acquisition unit 304 acquires the above-mentioned light receiving angle α1 and light receiving angle α2. In S1407, the derivation unit 305 fits the XYZ value histogram calculated by using the parameters of the model regarding the distribution characteristics of the bright material to the XYZ value histogram extracted as the color information corresponding to the region 1101. The derivation unit 305 uses a normal distribution as f (θ) as in the first embodiment. The parameters determined here are (μ1, σ1, M1).

In S1408, the calculation unit 1301 virtualizes the bright material corresponding to the region 1101 based on the distribution information representing the distribution characteristics of the bright material corresponding to the region 1101 and the condition information representing the geometric conditions corresponding to the region 1102. Calculate color information. Specifically, the calculation unit 1301 generates N random numbers of the rotation angle θ based on f (θ) shown in the equation (10).

Further, the following equations (11) and (12) are established.

The calculation unit 1301 calculates N wavelengths λ by substituting the generated N θs into the equation (12). The calculation unit 1301 calculates an XYZ value histogram based on the calculated wavelength λ and the color matching function, and normalizes the calculated XYZ value histogram by multiplying it by M1 / N.

In S1409, the evaluation unit 1302 calculates the similarity between the XYZ value histogram calculated as virtual color information in S1408 and the XYZ value histogram corresponding to the region 1102 extracted in S1405 as the evaluation value. As the degree of similarity, as described above, the area of the region where the histograms overlap, the Kullback-Leibler divergence, or the like can be used. In S1410, the output unit 306 outputs the evaluation value calculated by the evaluation unit 1302 to the display device 2.

<Effect of the second embodiment>
As described above, the image processing apparatus in the present embodiment calculates color information for two target areas having different observation conditions when the observation conditions of one target area are matched with the observation conditions of the other target area. do. Based on the calculated color information, the difference in the distribution characteristics of the bright material between the two regions is evaluated. This makes it possible to evaluate the difference in the distribution characteristics of the bright material between the two regions, such as between products and parts, with high accuracy.

<Modification example>
In the present embodiment, the virtual color information is calculated so that the observation condition of the region 1101 is matched with the observation condition of the region 1102, but both the observation condition of the region 1101 and the observation condition of the region 1102 are matched with the target observation condition. The virtual color information of each target area may be calculated as described above. In this case, in S1407, the derivation unit 305 derives the distribution characteristics of the bright material for each of the region 1101 and the region 1102. In S1408, the calculation unit 1301 calculates virtual color information when each target area is imaged under the target observation conditions. The target observation condition may be acquired by the acquisition unit 304 as condition information in S1406, for example. In S1409, the evaluation unit 1302 calculates the similarity between the virtual color information of the area 1101 and the virtual color information of the area 1102 as an evaluation value. This makes it possible to evaluate the difference in the distribution characteristics of the bright material between the two regions under arbitrary observation conditions.

Further, in the present embodiment, the color information of the bright material is compared in order to evaluate the difference in the distribution characteristic of the bright material between the two regions, but the distribution characteristic of the bright material is not calculated. You may compare each other directly. In this case, in S1407, the derivation unit 305 derives the distribution characteristics of the bright material for each of the region 1101 and the region 1102. The processing of S1408 is not performed, and in S1409, the evaluation unit 1302 evaluates the difference between the distribution information (μ1, σ1, M1) corresponding to the region 1101 and the distribution information (μ2, σ2, M2) corresponding to the region 1102. Calculated as. Specifically, the evaluation unit 1302 calculates | σ1-σ2 |, | μ1-μ2 |, | M1-M2 | as evaluation values. Thereby, even if the image processing apparatus 1 does not have the calculation unit 1301 for calculating the virtual color information, it is possible to evaluate the difference in the distribution characteristics of the bright material between the two regions.

Further, in the present embodiment, the difference in the distribution characteristics of the bright material between the two regions is evaluated, but the number of the target regions is not limited to two, and may be three or more.

[Other embodiments]
The present invention supplies a program that realizes one or more functions of the above-described embodiment to a system or device via a network or storage medium, and one or more processors in the computer of the system or device reads and executes the program. It can also be realized by the processing to be performed. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

1 Image processing device 301 Input unit 303 Extraction unit 304 Acquisition unit 305 Derivation unit

Claims (10)

An input means for inputting image data obtained by imaging an object containing a bright material,
An acquisition means for acquiring condition information representing geometric conditions when imaging the object, and
An extraction means for extracting color information of the bright material in the target region of the object based on the image data, and an extraction means.
A derivation means for deriving the distribution characteristics of the bright material in the target region based on the condition information and the color information of the bright material.
An image processing device characterized by having. The image processing apparatus according to claim 1, wherein the bright material is a bright material whose color changes according to observation conditions due to thin film interference. The image processing apparatus according to claim 1 or 2, wherein the extraction means extracts a histogram corresponding to the color of a bright spot in an image represented by the image data as color information of the bright material. The derivation means according to claim 3, wherein the derivation means derives the distribution characteristics of the brilliant material by fitting the histogram with a parameter for specifying the distribution characteristics of the brilliant material. Image processing device. The image processing apparatus according to claim 4, wherein the parameter includes the orientation of the bright material in the target region. The image processing apparatus according to claim 5, wherein the parameters are the orientation of the bright material and the number of the bright materials included in the target area. The image processing apparatus according to any one of claims 3 to 6, wherein the extraction means detects the bright spot by binarizing the pixel value of the image. The acquisition means acquires the condition information corresponding to each of the plurality of target areas, and obtains the condition information.
The derivation means derives the distribution characteristics of the bright material for each of the plurality of target regions.
Claims 1 to claim further include an evaluation means for evaluating a difference in the distribution characteristics of the bright material among a plurality of the target regions based on the condition information and the distribution characteristics of the bright material. 7. The image processing apparatus according to any one of 7. A program for making a computer function as each means of the image processing apparatus according to any one of claims 1 to 8. An input step for inputting image data obtained by imaging an object containing a bright material,
An acquisition step for acquiring condition information representing geometric conditions when imaging the object, and
An extraction step of extracting color information of the bright material in the target region of the object based on the image data, and an extraction step.
A derivation step for deriving the distribution characteristics of the bright material in the target region based on the condition information and the color information of the bright material, and
An image processing method characterized by having.

Images