JP2021196231A - Evaluation index calculation device, evaluation index calculation method, and evaluation index calculation program - Google Patents

Abstract

To provide an evaluation index calculation device, an evaluation index calculation method, and an evaluation index calculation program, in which an evaluation index representing an optical state of dimming sheet may be diversified.SOLUTION: Disclosed is an evaluation index calculation device for calculating an evaluation index representing an optical state of a dimming sheet 10, in which a photographing object 13 in a predetermined environment acquires an evaluation image PE photographed via the dimming sheet 10, and which calculates the blur degree of the evaluation image PE by the dimming sheet 10 as an evaluation index using the image analysis of the evaluation image PE. The image analysis is the frequency analysis of the evaluation image PE, and the degree of blur is a magnitude of a frequency component in a predetermined frequency band.SELECTED DRAWING: Figure 1

Description

The present invention relates to an evaluation index calculation device for calculating an evaluation index of an optical state of a dimming sheet, an evaluation index calculation method, and an evaluation index calculation program.

The visibility of an object through a dimming sheet having liquid crystal molecules between the transparent electrodes changes according to the voltage change between the transparent electrodes. The normal type dimming sheet that keeps opaque when not energized is applied to digital signage screens and store exteriors that are required to be opaque for a long time. The reverse type dimming sheet that keeps transparency when not energized is applied to glass partitions and windows of moving objects that are required to be transparent in an emergency.

Stabilization of the optical state of the light control sheet and improvement of reproducibility will expand the use of the light control sheet in various fields and accelerate the spread of the light control sheet. The haze conforming to JIS K 7136: 2000 is adopted as an evaluation index of the optical state of the dimming sheet. For example, setting the haze of the dimming sheet to a predetermined range of 95% or more as an opaque control item and setting the haze of the dimming sheet to a predetermined range of 12% or less as a transparent control item. Has been proposed (see, for example, Patent Document 1).

Japanese Patent No. 6493598

By the way, a transparent body such as a glass plate or an acrylic plate to which a dimming sheet is attached is installed in a building or a moving body. Being able to evaluate the optical state of a dimming sheet in a state where it is installed in a building or a moving body can further accelerate the spread of the dimming sheet. However, haze measurement, which measures the amount of light in a dark place by arranging a dedicated optical device such as an integrating sphere, makes it rather troublesome to evaluate the optical state in buildings and moving objects, so a dimming sheet is used. In the field of use, a new evaluation index capable of evaluating the optical state of the dimming sheet even when the dimming sheet is installed is eagerly desired.

The request for a new evaluation index indicating the optical state of the dimming sheet is not limited to the device of the dimming sheet, but also at the time of development of the dimming sheet, at the time of shipping the dimming sheet, at the time of monitoring the dimming sheet, etc. It is common in various situations where the dimming sheet is handled.

An object of the present invention is to provide an evaluation index calculation device, an evaluation index calculation method, and an evaluation index calculation program that enable diversification of evaluation indexes representing the optical state of a dimming sheet.

The evaluation index calculation device for solving the above problems is an evaluation index calculation device that calculates an evaluation index representing the optical state of the dimming sheet, and is in a predetermined environment photographed through the dimming sheet. An image relating to a subject is defined as an evaluation image, and a calculation unit is provided for calculating the degree of blurring of the evaluation image by the dimming sheet as the evaluation index by using image analysis of the evaluation image.

The evaluation index calculation method for solving the above-mentioned problems is an evaluation index calculation method for calculating an evaluation index representing an optical state of a dimming sheet, and is in a predetermined environment photographed through the dimming sheet. An image relating to a subject is defined as an evaluation image, and the degree of blurring of the evaluation image by the dimming sheet is calculated as the evaluation index by using image analysis of the evaluation image.

The evaluation index calculation program for solving the above problems relates to a subject in a predetermined environment photographed via the dimming sheet by an evaluation index calculation device that calculates an evaluation index representing the optical state of the dimming sheet. An image is defined as an evaluation image, and the degree of blurring of the evaluation image by the dimming sheet is calculated as the evaluation index by using image analysis of the evaluation image.

According to each of the above configurations, the degree of blurring of the evaluation image taken through the light control sheet by the light control sheet is calculated by using the image analysis of the evaluation image. The shooting of the evaluation image is realized not by a dedicated optical device such as an integrating sphere, but by a general-purpose shooting device such as a smartphone or a tablet terminal. The degree of blurring of the evaluation image is calculated by using various image analyzes such as frequency analysis and edge strength analysis of the evaluation image. The degree of blurring by the dimming sheet is the degree of blurring of the evaluation image taken without passing through the dimming sheet, the degree of blurring of the evaluation image taken with the dimming sheet transparent, and a predetermined value estimated to correspond to these. It can be calculated based on the value. The degree of blurring by the dimming sheet tends to increase as the haze of the dimming sheet increases, and it tends to change in synchronization with the haze change of the dimming sheet. For example, it is possible to diversify the evaluation index showing the optical state of the dimming sheet.

The evaluation index calculation device uses the degree of blurring when the optical state of the light control sheet is normal and the degree of blurring calculated by the evaluation index calculation device to determine whether the light control sheet is normal. It may have a determination unit for determining whether or not.

According to the evaluation index calculation device, it is possible to determine whether or not the dimming sheet is normal by using the new evaluation index.

In the evaluation index calculation device, the image analysis may be a frequency analysis of the evaluation image.
The sharpness of repetition in a short cycle in the evaluation image is detected as the magnitude of the frequency component in the high frequency band. On the other hand, the sharpness of repetition in a long period in the evaluation image is detected as the magnitude of the frequency component in the low frequency band. In this regard, since the degree of blurring is calculated through frequency analysis of the evaluation image in the evaluation index calculation device, it is possible to express the optical state of the light control sheet in detail by the degree of blurring.

In the evaluation index calculation device, the degree of blurring may be the magnitude of a frequency component in a predetermined frequency band.
According to the evaluation index calculation device, the magnitude of the frequency component in a predetermined frequency band is calculated as the degree of blurring, and therefore, for example, the sharpness of repetition in a short cycle or the sharpness of repetition in a long cycle is special. It is also possible to provide a standardized evaluation index.

In the evaluation index calculation device, the magnitude of the frequency component may be the magnitude of the radial distribution obtained from the Fourier transform of the image.
According to the above configuration, since the magnitude of the frequency component in the radial distribution is calculated as the degree of blurring, whether the spatial frequency component appears in the one-dimensional direction or the spatial frequency component appears in the two-dimensional direction. , The frequency component of the size according to the optical state of the dimming sheet can be calculated. That is, the restrictions related to spatial anisotropy are alleviated in the object to be photographed. Therefore, it is possible to expand the situations in which the evaluation index calculation device is applied in various fields, and to increase the versatility of the evaluation index calculation device.

In the evaluation index calculation device, the predetermined frequency band may be a band higher than the frequency at which the energy is maximum in the radial distribution obtained from the Fourier transform of the image.

As described above, the sharpness of repetition in a short cycle in the evaluation image is detected as the magnitude of the frequency component in the high frequency band. On the other hand, the frequency component in the low frequency band tends to show only the existence or nonexistence of the object through the dimming sheet, and tends to show the maximum energy in the radial distribution. In this regard, with the evaluation index calculation device described above, the degree of blurring of the evaluation image is calculated in a band higher than the frequency at which the energy is maximum, so that the outline of the object is not the rough visibility of the existence of the object. It is also possible to evaluate whether or not the dimming sheet provides fine visibility.

In the evaluation index calculation device, the magnitude of the frequency component when the optical state of the dimming sheet is normal is stored as a normal range, and when the magnitude of the frequency component is within the normal range, the adjustment is performed. It may be determined that the optical state of the optical sheet is normal.

According to the evaluation index calculation device, it is possible to determine that the optical state of the light control sheet is normal from the size of the frequency component by image analysis.
In the evaluation index calculation device, the degree of blurring of the evaluation image when the driving voltage is not applied to the dimming sheet is determined by the degree of blurring of the evaluation image when the driving voltage is applied to the dimming sheet. A standardized value may be calculated.

According to the evaluation index calculation device, the optical state of the dimming sheet when the driving voltage is not applied is calculated as the evaluation index based on the state of the dimming sheet when the driving voltage is applied. Therefore, it is possible to evaluate the optical state of the dimming sheet, including whether or not the driving device for driving the dimming sheet is normal.

The evaluation index calculation device, the evaluation index calculation method, and the evaluation index calculation program according to the present invention enable diversification of the evaluation index representing the optical state of the dimming sheet.

A block diagram showing the configuration of the evaluation index calculation device together with the dimming sheet. A block diagram functionally showing the configuration of the image analysis unit. A power spectrum diagram showing an example of sampling of a sample group for calculating an evaluation index. A flowchart showing a process flow in the evaluation index calculation method. FIG. 3 is a device configuration diagram showing a device for photographing an evaluation image in device example 1. FIG. 3 is a device configuration diagram showing a device for photographing an evaluation image in device example 2. (A) Evaluation image example 1-1 and (b) Evaluation image example 2-1 are shown. (A) Evaluation image example 1-2 and (b) Evaluation image example 2-2 are shown. (A) Evaluation image example 1-3 and (b) Evaluation image example 2-3 are shown. (A) shows the FFT conversion image of the evaluation image example 1-1, and (b) shows the FFT conversion image of the evaluation image example 2-1. (A) shows the FFT conversion image of the evaluation image example 1-2, and (b) shows the FFT conversion image of the evaluation image example 2-2. (A) shows the FFT conversion image of the evaluation image example 1-3, and (b) shows the FFT conversion image of the evaluation image example 2-3. (A) shows the evaluation image example 3-1 and (b) shows the evaluation image example 4-1. (A) shows the evaluation image example 3-2, and (b) shows the evaluation image example 4-2. (A) shows the evaluation image example 3-3, and (b) shows the evaluation image example 4-3. (A) shows the FFT conversion image of the evaluation image example 3-1 and (b) shows the FFT conversion image of the evaluation image example 4-1. (A) shows the FFT conversion image of the evaluation image example 3-2, and (b) shows the FFT conversion image of the evaluation image example 4-2. (A) shows the FFT conversion image of the evaluation image example 3-3, and (b) shows the FFT conversion image of the evaluation image example 4-3. The graph which shows the radial direction distribution of the test example 1 obtained from the FFT conversion image of the evaluation image example 1-1, 1-2, 1-3. The graph which shows the angular direction distribution of the test example 1 obtained from the FFT conversion image of the evaluation image example 1-1, 1-2, 1-3. The graph which shows the radial direction distribution of the test example 3 obtained from the FFT conversion image of the evaluation image example 3-1, 3-2, 3-3. The graph which shows the angular direction distribution of the test example 3 obtained from the FFT conversion image of the evaluation image example 3-1, 3-2, 3-3. The graph which shows the radial direction distribution of the test example 2 obtained from the FFT conversion image of the evaluation image example 2-1, 2-2, 2-3. The graph which shows the angular direction distribution of the test example 2 obtained from the FFT conversion image of the evaluation image example 2-1, 2-2, 2-3. The graph which shows the radial direction distribution of the test example 4 obtained from the FFT conversion image of the evaluation image example 4-1, 4-2, 4-3. The graph which shows the angular direction distribution of the test example 4 obtained from the FFT conversion image of the evaluation image example 4-1, 4-2, 4-3. A graph showing the relationship between the standard value of the sample integrated value obtained from the radial distribution of each test example and the drive voltage together with the haze.

An evaluation index calculation device, an evaluation index calculation method, and an evaluation index calculation program will be described with reference to FIGS. 1 to 25. First, the configuration of the dimming sheet to be evaluated will be described, and then the evaluation index calculation device, the evaluation index calculation method, and the evaluation index calculation program will be described.

[Dimming sheet 10]
The light control sheet 10 is configured so that the light transmittance can be changed. The light transmittance of the light control sheet 10 is changed by changing the drive voltage applied to the light control sheet 10. The type of the light control sheet 10 includes an electrochromic sheet in addition to the liquid crystal light control sheet. The transmittance of the liquid crystal dimming sheet is controlled by controlling the orientation of the liquid crystal molecules by applying the driving voltage V. The transmittance of the electrochromic sheet is controlled by controlling the charge of the electrochromic material by applying the drive voltage V.

The dimming sheet 10 projects windows provided in various buildings such as commercial facilities and public facilities, partitions installed in offices and medical facilities, show windows installed in exhibition facilities and cultural facilities, and images. It is attached to various transparent bodies 11 such as a base material of a screen and a window provided in a moving body such as a vehicle or an aircraft. The light control sheet 10 is attached to the surface of a transparent body 11 such as a window glass, or sandwiched between two transparent bodies such as laminated glass.

The shape of the dimming sheet 10 is not only a flat surface that follows the outer shape of the transparent body 11, but also a two-dimensional curved surface such as a cylinder or an ellipsoid, and a three-dimensional curved surface such as a spherical surface or an ellipsoid. , And indefinite shapes other than geometric shapes. The color of the dimming sheet 10 in a state of having high transmittance is colorless and transparent or colored transparent. The color of the dimming sheet 10 in a state of having low transmittance is achromatic color or chromatic color.

The model of the dimming sheet 10 is a normal type or a reverse type. The normal type dimming sheet has a relatively high light transmittance when the dimming sheet 10 is energized, while it has a relatively low light transmittance when the dimming sheet 10 is not energized. The reverse type dimming sheet has a relatively low light transmittance when the dimming sheet 10 is energized, while it has a relatively high light transmittance when the dimming sheet 10 is not energized.

The light control sheet 10 includes a first transparent electrode, a second transparent electrode, and a light control layer that extend in the plane direction of the light control sheet 10. In the light control sheet 10, the first transparent electrode, the light control layer, and the second transparent electrode are laminated in this order, and the light control layer is sandwiched between the first transparent electrode and the second transparent electrode. The first transparent electrode and the second transparent electrode are electrically connected to the dimming control device 30.

The first transparent electrode and the second transparent electrode have light transmission through which visible light is transmitted. The light transmittance of the first transparent electrode enables visual recognition of an object through the dimming sheet 10. The light transmittance of the second transparent electrode also enables visual recognition of an object through the dimming sheet 10. The material constituting each transparent electrode is, for example, any one selected from the group consisting of indium tin oxide, fluorine-doped tin oxide, tin oxide, zinc oxide, carbon nanotubes, and poly (3,4-ethylenedioxythiophene). Is.

The light control layer included in the liquid crystal light control sheet contains a liquid crystal composition. Examples of liquid crystal molecules contained in the liquid crystal composition include Schiff base type, azo type, azoxi type, biphenyl type, terphenyl type, benzoic acid ester type, trans type, pyrimidine type, cyclohexanecarboxylic acid ester type, phenylcyclohexane type, and the like. It is at least one selected from the group consisting of dioxane systems. The holding type of the liquid crystal composition is a polymer dispersion type or a capsule type. The polymer-dispersed type holds the liquid crystal composition in a large number of voids dispersed in the polymer layer. The polymer-dispersed type includes a polymer network type having a polymer network having a three-dimensional network. The polymer network type holds the liquid crystal composition in the mesh-like voids. The capsule type holds the liquid crystal composition having a capsule shape in the polymer layer.

The reverse type liquid crystal light control sheet further includes an alignment film between the light control layer and the first transparent electrode, and between the light control layer and the second transparent electrode. The material constituting the alignment film is, for example, an organic compound such as polyimide, polyamide, polyvinyl alcohol, and a cyanide compound, an inorganic compound such as silicone, silicon oxide, and zirconium oxide, or a mixture thereof.

The alignment film is, for example, a vertical alignment film or a horizontal alignment film. The vertical alignment film is oriented in the major axis direction of the liquid crystal molecules so as to be perpendicular to the surface opposite to the surface in contact with the first transparent electrode and the surface opposite to the surface in contact with the second transparent electrode. Align. The horizontal alignment film is in the major axis direction of the liquid crystal molecules so as to be substantially parallel to the surface opposite to the surface in contact with the first transparent electrode and the surface opposite to the surface in contact with the second transparent electrode. To orient.

The dimming layer included in the electrochromic sheet contains an electrochromic material and an electrolyte. An example of an electrochromic material is one selected from organic compounds such as polyaniline derivatives, viologens, metallic phthalocyanines and phenanthroline complexes, and inorganic compounds such as tungsten trioxide and indium dioxide. An example of the electrolyte is a liquid electrolyte such as a lithium salt or a potassium salt, or a polymer solid electrolyte.

[Dimming system]
As shown in FIG. 1, the dimming system includes a dimming sheet 10, a photographing unit 12, an index calculation device 20 which is an example of an evaluation index calculation device, and a dimming control device 30 which is an example of a drive device.

The dimming control device 30 receives a drive command NS from the index calculation device 20 or receives an input signal MC of the operation unit to drive the dimming sheet 10 for calculating the evaluation index Qe. The photographing unit 12 photographs the photographing object 13 via the dimming sheet 10 to generate an evaluation image PE of the photographing object 13. The photographing unit 12 transmits the evaluation image PE generated by the photographing unit 12 to the index calculation device 20. The index calculation device 20 calculates the degree of blurring of the evaluation image PE by the dimming sheet 10 by using the image analysis of the evaluation image PE.

The degree of blurring calculated by the index calculation device 20 is the evaluation index Qe of the dimming sheet 10. The degree of blurring by the light control sheet 10 is a value indicating whether or not the light control sheet 10 has an optical state, which is an evaluation item. The optical state, which is an evaluation item, is, for example, having a haze of a predetermined value or more, having a clarity of a predetermined value or less, and the like.

For example, the degree of blurring by the light control sheet 10 is a value indicating whether or not the light control sheet 10 has a high haze such that the presence or absence of an object is not visually recognized through the light control sheet 10. On the contrary, the degree of blurring by the dimming sheet 10 is a value indicating whether or not the dimming sheet 10 has a haze low enough to visually recognize the existence or nonexistence of the object through the dimming sheet 10.

For example, the degree of blurring by the light control sheet 10 is a value indicating whether or not the light control sheet 10 has a high haze such that the contour of an object is not visually recognized through the light control sheet 10. On the contrary, the degree of blurring by the light control sheet 10 is a value indicating whether or not the light control sheet 10 has a low haze such that the contour of the object is visually recognized through the light control sheet 10.

For example, the degree of blurring by the dimming sheet 10 is a value indicating whether or not the dimming sheet 10 has a low clarity such that the type of the object is not visually recognized through the dimming sheet 10. On the contrary, the degree of blurring by the dimming sheet 10 is a value indicating whether or not the dimming sheet 10 has a high degree of clarity such that the details of the object can be visually recognized through the dimming sheet 10.

The index calculation device 20 is, for example, when the evaluation index Qe calculated by the index calculation device 20 and the optical state of the light control sheet 10 are normal during the manufacturing, construction, or use period of the light control sheet 10. It is determined whether or not the dimming sheet 10 is normal by comparing it with the index threshold value Qm which is an evaluation index of. That is, the evaluation index Qe calculated by the index calculation device 20 is used for determining the abnormality of the dimming sheet 10.

The index calculation device 20 specifies, for example, a drive voltage V estimated to have a normal optical state in the dimming sheet 10 during manufacturing, construction, or use of the dimming sheet 10. The dimming control device 30 applies the specified drive voltage V to the dimming sheet 10, for example, by receiving a drive command NS from the index calculation device 20 or receiving an input signal MC by the user. That is, the evaluation index Qe calculated by the index calculation device 20 is used for drive correction of the dimming sheet 10.

For example, in the process of determining deterioration, the index calculation device 20 stores the previous index Q0, which is the previously calculated evaluation index, during the period of use of the dimming sheet 10, and the evaluation index Qe and the previous index calculated this time. It is determined whether or not the deviation from Q0 is within a predetermined range. Then, the index calculation device 20 determines that the dimming sheet 10 has not deteriorated over time based on the difference between the evaluation index Qe calculated this time and the previous index Q0 within a predetermined range. On the other hand, the index calculation device 20 determines that the dimming sheet 10 has deteriorated over time based on the fact that the deviation between the evaluation index Qe calculated this time and the previous index Q0 is out of the predetermined range. That is, the evaluation index Qe calculated by the index calculation device 20 is used for determining the deterioration of the dimming sheet 10.

The photographing unit 12 and the index calculation device 20 are configured to enable transmission of the evaluation image PE by the photographing unit 12 and reception of the evaluation image PE by the index calculation device 20 through wired communication or wireless communication. The index calculation device 20 may include, for example, a maintenance device that is communicably connected to the photographing unit 12 without a network, a server device that is communicably connected to the photographing unit 12 via a network, or a photographing unit 12. It is a device built into the installed user terminal. The index calculation device 20 is, for example, a server device that receives a calculation application for the evaluation index Qe from a user terminal equipped with a photographing unit 12 and receives an evaluation image PE photographed via the dimming sheet 10 from the user terminal. Is.

The index calculation device 20 and the dimming control device 30 can transmit a drive command NS by the index calculation device 20 and receive a drive command NS by the dimming control device 30, for example, through wired communication or wireless communication. It is configured. The index calculation device 20 is, for example, a maintenance device that is communicably connected to the dimming control device 30 without going through a network, a server device that is communicably connected to the dimming control device 30 via a network, or It is an electronic device provided with hardware common to a part of the dimming control device 30.

The photographing unit 12 receives a two-dimensional image sensor in which photographing elements such as CCD (Charged coupled devices) or CMOS (Complementary metal oxide semiconductor) are arranged in two dimensions, and the light emitted from the photographing target 13 is received by the image sensor. It is equipped with a photographing optical system that forms an image on a surface. The photographing unit 12 forms an image of the light emitted from the photographing object 13 on the light receiving surface of the image sensor, and records the still image of the photographing object 13 through the dimming sheet 10 as the evaluation image PE which is the image data. The photographing unit 12 is, for example, a camera mounted on a smartphone or a tablet terminal, a surveillance camera installed in various facilities such as commercial facilities and public facilities, and a general camera such as a digital camera.

The photographing unit 12 generates the evaluation image PE. The evaluation image PE is an image in which the photographing target 13 in a predetermined environment is photographed through the dimming sheet 10. The shooting of the shooting target 13 is performed through the dimming sheet 10 with the shooting unit 12 and the shooting target 13 placed in a predetermined environment.

The predetermined environment for shooting the evaluation image PE is the setting environment of the shooting unit 12, the lighting environment for the shooting unit 12, and the lighting environment for the shooting target 13. The setting environment of the shooting unit 12 is, for example, to set the manual focus distance at the time of shooting to a predetermined value, and to set the aperture value and the shutter speed to a predetermined value. The lighting environment for the shooting unit 12 is such that, for example, the lighting is turned on at the place where the shooting unit 12 is placed, or the amount of solar radiation at the place where the shooting unit 12 is placed is estimated to be a predetermined amount. be. The lighting environment for the shooting target 13 is, for example, that the lighting is turned on at the place where the shooting target 13 is placed, or the amount of solar radiation at the place where the shooting target 13 is placed is estimated to be a predetermined amount. be. Estimating that the amount of solar radiation is a predetermined amount means that, for example, the shooting date and time is a predetermined date and time, and the type of weather at the time of shooting is fine. As described above, the shooting of the evaluation image PE is realized not by a dedicated optical device such as an integrating sphere but by a general-purpose shooting device such as a smartphone or a tablet terminal.

The shooting target 13 is a shooting target used for calculating the evaluation index Qe. The object 13 to be photographed is a test pattern such as a printed matter prepared in advance, a person image, an artificial object such as a building or a building visually recognized through the dimming sheet 10, and a natural landscape visually recognized through the dimming sheet 10.

A test pattern is a printed matter used to evaluate physical performance such as contrast, resolution, noise, etc. by the degree of separation between black and white. The test pattern is, for example, a pattern in which white and black thin lines are arranged in parallel alternately, a pattern in which white and black thin lines are arranged radially, and white and black thin lines having different line widths are alternately arranged. Patterns arranged in parallel, patterns in which white and black fine lines having different linear densities are arranged, and the like. The test pattern is, for example, characters or alphanumericals having different sizes from each other. In addition, the test pattern is, for example, a test chart containing text such as characters and alphanumericals used for resolution measurement of digital cameras, a resolution chart compliant with ISO12233-2000, and JIS X 0527 used for performance evaluation of barcode readers. Compliant barcode test charts, etc.

[Index calculation device 20]
The index calculation device 20 is not limited to processing all kinds of processing by software. For example, the index calculation device 20 includes dedicated hardware such as an integrated circuit for a specific application (Graphics Processing Unit: GPU, Application Specific Integrated Circuit: ASIC) that executes at least a part of various processes. May be good. The index calculation device 20 is also configured as a circuit including one or more dedicated hardware circuits such as an ASIC, one or more processors operating according to a computer program, or a combination thereof. In the following, an example will be described in which the index calculation device 20 stores the evaluation index calculation program in a readable medium, reads out the evaluation index calculation program stored in the readable medium, executes the program, and performs various processes.

The index calculation device 20 includes a control unit 21, an image analysis unit 22, a storage unit 23, and a communication unit 24.
The storage unit 23 stores the evaluation index calculation program 23A for calculating the evaluation index and the index calculation information 23B used for executing the evaluation index calculation program. The index calculation information 23B includes an evaluation voltage V0 and sampling conditions. The evaluation voltage V0 is a drive voltage V applied to the dimming sheet 10, and is used as an initial value applied for calculating the evaluation index Qe. The sample extraction condition is a condition for extracting the sample group used for the calculation of the evaluation index Qe from the distribution of the feature vector. The sampling conditions are set so that a sample group suitable for evaluation is extracted based on the results of tests conducted in advance.

For example, when it is an evaluation item of the optical state to have a high haze such that the presence or absence of an object cannot be visually recognized through the dimming sheet 10, a sample group suitable for evaluating whether or not the object has the high haze is selected. The conditions for extracting from the distribution of the feature vector are stored as sampling conditions. On the contrary, when the evaluation item is that the dimming sheet 10 has a low haze that allows the presence or absence of an object to be visually recognized through the dimming sheet 10, it is suitable for evaluating whether or not the dimming sheet 10 has the low haze. The conditions for extracting the sample group from the distribution of the feature vector are stored as the sample extraction conditions.

For example, when it is an evaluation item of the optical state to have a high haze such that the contour of the object is not visually recognized through the dimming sheet 10, a sample group suitable for evaluating whether or not the object has the high haze is selected. The conditions for extracting from the distribution of the feature vector are stored as sampling conditions. On the contrary, when it is an evaluation item that the dimming sheet 10 has a low haze such that the details of the object can be visually recognized through the dimming sheet 10, it is suitable for evaluating whether or not the dimming sheet 10 has the low haze. The conditions for extracting the sample group from the distribution of the feature vector are stored as the sample extraction conditions.

For example, when it is an evaluation item to have a low clarity such that the contour of an object is not visually recognized through the dimming sheet 10, a sample group suitable for evaluating whether or not the object has the low clarity is used as a feature vector. The conditions for extracting from the distribution are stored as sampling conditions. On the contrary, when it is an evaluation item that the dimming sheet 10 has a high clarity such that the details of the object can be visually recognized through the dimming sheet 10, it is suitable for evaluating whether or not the dimming sheet 10 has the high clarity. The conditions for extracting the sample group from the distribution of the feature vector are stored as the sample extraction conditions.

The index calculation information 23B may include data required for various processes using the evaluation index Qe, such as the index threshold value Qm, the upper limit voltage Vmax, and the step voltage Vs. The index threshold value Qm is a threshold value of the evaluation index Qe obtained through the dimming sheet 10 satisfying the evaluation items, and is determined based on a test or the like conducted in advance. The upper limit voltage Vmax is the maximum value of the drive voltage applied to the dimming sheet 10 by the dimming control device 30. The step voltage Vs is a change amount of the drive voltage V that can be changed for each opportunity to calculate the evaluation index Qe in the drive correction.

The control unit 21 reads the evaluation index calculation program 23A from the storage unit 23, interprets the evaluation index calculation program 23A, and causes the image analysis unit 22, the storage unit 23, and the communication unit 24 to execute various processes. For example, the control unit 21 causes the communication unit 24 to acquire the evaluation image PE from the photographing unit 12, and causes the image analysis unit 22 to perform image analysis using the evaluation image PE acquired by the communication unit 24. For example, the control unit 21 causes the communication unit 24 to execute an output process for outputting the evaluation index Qe to the outside by using the result of the image analysis by the image analysis unit 22. For example, the control unit 21 is an example of a determination unit, and uses the result of image analysis by the image analysis unit 22 to execute various processes such as abnormality determination of the dimming sheet 10 and drive correction of the dimming sheet 10. ..

[Image analysis unit 22]
The image analysis unit 22 executes image analysis of the evaluation image PE, and calculates the degree of blurring of the evaluation image PE by the dimming sheet 10 as the evaluation index Qe. The image analysis unit 22 is an example of a calculation unit.

As shown in FIG. 2, the image analysis unit 22 functions as a grayscale processing unit 22A, an FFT processing unit 22B, and a frequency analysis unit 22C. The gray scale processing 22A performs (i) gray scale conversion of the evaluation image PE. The FFT processing unit 22B performs (ii) a two-dimensional fast Fourier transform using the grayscaled image data. The frequency analysis unit 22C executes frequency analysis of the (iii) Fourier transform image using the processing result of the FFT processing unit 22B.

In the image analysis performed by the image analysis unit 22, the degree of blurring of the image indicates the degree of clarity of the evaluation image PE, that is, the optical density such as the brightness of the evaluation image PE and the optical density such as the brightness of the background image. It may be to calculate the difference between. Alternatively, the image analysis performed by the image analysis unit 22 may be to calculate the optical density of the pixel having the pixel value corresponding to the edge intensity in the contour of the evaluation image PE as the degree of blurring of the image. For example, it is also possible to determine each pixel corresponding to the contour of the evaluation image by using the average value of the edge strength obtained by the edge detection process based on the luminance value of the pixel.

(i) Grayscale conversion changes the image data in RGB space to the image data in YUV space. Grayscale is a process of performing an intermediate value method that outputs an intermediate value using RGB values as a result, a weighted averaging method that outputs an average value of RGB values weighted by each different coefficient as a result, and the like. Further, the gray scale may be a process in which a weighted averaging method and a correction are combined to output a value obtained by applying gamma correction to the weighted average value of the RGB values weighted by each different coefficient as a result. Further, the grayscale may be a process of performing a simple averaging method that outputs a simple average value of RGB values as a result, a median method that outputs a median value among RGB values as a result, and the like.

(Ii) The two-dimensional fast Fourier transform is performed using the grayscaled image data, and the distribution of the feature vectors in the two-dimensional Fourier transform image is calculated. The feature vector is at least one of a radial vector and an angular vector. The radial distribution p (r) is an example of a frequency component, and is the sum of power spectra on a concentric region existing at a distance r from the center of the two-dimensional Fourier transformed image. The angular distribution q (θ) is the sum of the power spectra on the linear region of the angle θ with respect to the horizontal axis. The distribution of the feature vector is determined by what spatial frequency the wave with is strongly included in the image data after grayscale, and in what direction the strong wave is included, that is, the radial direction and the angle. It shows the periodicity of shading in the image in the direction.

(Iii) In the frequency analysis, a sample group satisfying the sample extraction condition is extracted from the distribution of the feature vector. In the frequency analysis, the evaluation index Qe is calculated from the feature vector of the extracted sample group. As described above, the sampling condition is a condition for extracting the sample group used for calculating the evaluation index from the distribution of the feature vector. In the frequency analysis, for example, the sum of the feature vectors of all the samples satisfying the sampling conditions is calculated as the sample integration value SumV. In the frequency analysis, the evaluation index Qe is calculated using the calculated sample integrated value SumV and the reference value for standardizing the sample integrated value SumV.

[Sampling]
Next, an example of extracting a sample group in the above frequency analysis will be described below. FIG. 3 shows the sample group extracted in the power spectrum of the Fourier transformed image with dots.

As shown in FIG. 3, for example, when the distribution of the feature vector is the distribution in the radial direction, one example of the sampling condition is that the distance r is within a predetermined range, that is, the frequency is in a predetermined frequency band. Is. When it is an evaluation item that the contour of the object has a high haze that cannot be visually recognized, one example of the sampling condition is that the distance r is equal to or larger than the sampling threshold value rx. An example of the sample threshold value rx satisfies the following conditions 1 and 2 in the radial distribution.

(Condition 1)
Specimen threshold rx> Extraction reference distance rs
(Condition 2)
Sample threshold energy p (rx) <extraction reference energy p (rs) x 10%
Here, the extraction reference energy (rs) is, for example, the maximum value p (r) max indicating the largest energy in the radial direction distribution p (r). The extraction reference distance rs is a distance r that gives the maximum value p (r) max.

In the grayscaled image, the light and shade wave corresponding to the maximum value p (r) max is likely to be a wave indicating the existence or nonexistence of the object. On the other hand, in the grayscaled image, the light and shade wave corresponding to the energy smaller than the maximum value p (r) max is likely to be a wave showing the contour of an object or the like. Then, the extraction of the sample group based on the above conditions 1 and 2 excludes the waves that are likely to indicate the existence of the object from the sample group, and evaluates that the contour of the object is not visually recognized. A suitable sample group can be extracted.

As described above, the setting of the above condition 1 and the condition 2 excludes the low frequency wave below the shade wave corresponding to the feature vector by setting the extraction reference energy (rs) to the specific feature vector. It makes it possible to extract a sample group suitable for evaluating whether or not a high frequency wave is visually recognized. That is, the setting of the sampling condition is a condition that enables the calculation of the magnitude of the frequency component in the predetermined frequency band. In the frequency analysis, the sum of the feature vectors of all the samples extracted based on these sampling conditions is calculated as the sample integration value SumV (regions marked with dots in FIG. 3). Then, in the frequency analysis, the sample integration value SumV is divided by the reference value, and the evaluation index Qe standardized by the reference value is calculated.

When the evaluation item is to have a high haze such that the contour of the object is not visually recognized, the reference value is the sample integrated value of the evaluation image PE taken under a predetermined environment without passing through the dimming sheet 10. SumV, the sample integrated value SumV of the evaluation image PE taken with the dimming sheet 10 transparent, and a predetermined value estimated to correspond to these.

On the contrary, when the evaluation item is to have a haze low enough for the contour of the object to be visually recognized, the reference value is taken under a predetermined environment through a dimming sheet 10 to which the contour of the object is not visually recognized. It is a sample integrated value SumV of the evaluated evaluation image PE, or a predetermined value estimated to correspond to the sample integrated value SumV. In this case as well, the reference value may be separately calculated by the index calculation device 20 or may be stored in advance in the index calculation device 20.

When the optical state when the drive voltage V is applied to the dimming sheet 10 is an evaluation item, the reference value is, for example, the sample integration value when the drive voltage V is not applied to the dimming sheet 10. It is SumV or a predetermined value estimated to correspond to this.

On the contrary, when the optical state when the drive voltage V is not applied to the dimming sheet 10 is an evaluation item, the reference value is, for example, the sample integration when the drive voltage V is applied to the dimming sheet 10. It is a value SumV or a predetermined value estimated to correspond to the value SumV.

In any case, the reference value may be separately calculated by the index calculation device 20 or may be stored in advance in the index calculation device 20.
Further, when the distribution of the feature vector is the angular distribution q (θ), it is a sampling condition that the angle θ is within a predetermined range. Further, when the distribution of the feature vector is the angular distribution q (θ), the reference value is also taken from the angular distribution q (θ) of the evaluation image PE taken under a predetermined environment without passing through the dimming sheet 10. It is a sample integrated value obtained. When the distribution of the feature vector is the angular distribution q (θ), the reference value is the sample integrated value obtained from the angular distribution q (θ) of the evaluation image PE taken with the dimming sheet 10 transparent. Or, a predetermined value estimated to correspond to these.

The communication unit 24 executes to acquire the evaluation image PE from the photographing unit 12 in response to the command from the control unit 21. The communication unit 24 transmits, for example, a drive command NS required for calculating the evaluation index Qe to the dimming control device 30 in response to a command from the control unit 21, and notifies the external notification device of the result of the abnormality determination. It is possible to execute the driving command NS required for the driving correction of the dimming sheet 10 to the dimming control device 30. Further, the communication unit 24 may transmit the evaluation index Qe calculated by the image analysis unit 22 to the user terminal in response to a command from the control unit 21.

[Dimming control device 30]
Returning to FIG. 1, the dimming control device 30 is not limited to processing all kinds of processing by software. For example, the dimming control device 30 may include dedicated hardware such as an application specific integrated circuit (ASIC) that executes at least a part of various processes. The dimming control device 30 is also configured as a circuit including one or more dedicated hardware circuits such as an ASIC, one or more processors operating according to a computer program, or a combination thereof. In the following, an example will be described in which the dimming control device 30 stores a drive program in a readable medium, reads out the drive program stored in the readable medium, executes the drive program, and performs various processes.

The dimming control device 30 includes a control unit 31, a storage unit 32, and a drive unit 33. The storage unit 32 stores a drive program for driving the dimming sheet 10 and various data used for executing the drive program. The control unit 31 reads the drive program from the storage unit 32, interprets the drive program, and causes the drive unit 33 to generate the drive voltage V.

For example, the control unit 31 receives a drive command NS from the index calculation device 20 or receives an input signal MC of the operation unit, and causes the drive unit 33 to generate a drive voltage V for calculating the evaluation index Qe. The drive unit 33 applies the generated drive voltage V to the dimming sheet 10.

For example, the control unit 31 receives the drive command NS after the drive correction from the index calculation device 20, and stores the drive voltage V corrected by the drive correction in the storage unit 32. The control unit 31 receives the drive command NS from the index calculation device 20 or receives the input signal MC of the operation unit, reads the corrected drive voltage V from the storage unit 32, and drives the corrected drive voltage V. It is generated by the unit 33. The drive unit 33 applies the generated corrected drive voltage V to the dimming sheet 10.

[Index calculation method]
Next, the index calculation method executed by the index calculation device 20 will be described together with various processes such as normality determination, abnormality determination, and drive correction that can be executed by the dimming system. In the following, an example will be shown in which a reverse type liquid crystal dimming sheet is used, and whether or not the dimming sheet 10 has a high haze to the extent that the contour of an object is not visually recognized through the dimming sheet 10 is used as an evaluation item. ..

As shown in FIG. 4, the index calculation device 20 first transmits a drive command NS for generating the evaluation voltage V0 as the drive voltage V to the dimming control device 30. The dimming control device 30 generates an evaluation voltage V0 in response to the drive command NS, and applies an evaluation voltage V0 for making the dimming sheet 10 opaque to the dimming sheet 10. As a result, the dimming sheet 10 changes to an optical state for calculating the evaluation index Qe (step S11).

The index calculation device 20 may be configured to store a separate evaluation voltage V0 for each evaluation item. For example, as for the evaluation voltage V0, the evaluation voltage V0 corresponding to the evaluation item is selected by the index calculation device 20 according to the evaluation item selected by the user. For example, it is an evaluation item to have a high haze such that the presence or absence of an object is not visually recognized through the dimming sheet 10 and to have a high haze such that the contour of the object is not visually recognized through the dimming sheet 10. If so, the user selects one of the evaluation items. The index calculation device 20 reads out the evaluation voltage V0 corresponding to the evaluation item selected by the user, and transmits a drive command NS for setting the evaluation voltage V0 to the drive voltage V to the dimming control device 30.

Next, the photographing unit 12 photographs the evaluation image PE of the imaging target 13 via the light control sheet 10 to which the evaluation voltage V0 is applied. The index calculation device 20 acquires the evaluation image PE of the photographing target 13 photographed through the dimming sheet 10 from the photographing unit 12 (step S12).

Next, the index calculation device 20 executes image analysis of the acquired evaluation image PE, and calculates the degree of blurring of the evaluation image PE by the dimming sheet 10 as the evaluation index Qe. The image analysis performed by the index calculation device 20 is to calculate the degree of blurring of the evaluation image PE by the dimming sheet 10, and as described above, for example, grayscale, two-dimensional fast Fourier transform, and frequency analysis. (Step S13).

Next, the index calculation device 20 determines whether or not the calculated evaluation index Qe is equal to or greater than the index threshold value Qm. At this time, the index threshold value Qm is the lower limit value of the evaluation index Qe obtained in advance through the dimming sheet 10 having an appropriate optical state. The dimming sheet 10 having an appropriate optical state has a high haze such that the contour of the object is not visually recognized through the dimming sheet 10 (step S14).

When the index calculation device 20 determines that the evaluation index Qe is equal to or greater than the index threshold value Qm (YES in step S14), the index calculation device 20 stores the evaluation index Qe as the previous index Q0 (step S15). Then, the index calculation device 20 notifies the external notification device that the optical state of the dimming sheet 10 at the evaluation voltage V0 is normal (normality determination: step S16). In the index calculation device 20, in addition to the normality determination, or instead of comparing the evaluation index Qe and the index threshold value Qm in the normality determination, the deviation between the evaluation index Qe calculated this time and the previous index Q0 is within a predetermined range. It is also possible to perform the above-mentioned deterioration determination for determining whether or not the value is inside.

On the other hand, when the index calculation device 20 determines that the evaluation index Qe does not meet the index threshold value Qm (NO in step S14), the index calculation device 20 determines whether or not the current drive voltage V is the upper limit voltage Vmax (step S17). .. Next, when the index calculation device 20 determines that the current drive voltage V is not the upper limit voltage Vmax (NO in step S17), the index calculation device 20 generates a new drive voltage V by adding the step voltage Vs to the evaluation voltage V0. The drive command NS is transmitted to the dimming control device 30.

The dimming control device 30 generates a new drive voltage V in response to the drive command NS, and applies the generated drive voltage to the dimming sheet 10. As a result, the light transmittance of the dimming sheet 10 is further reduced. Then, by repeating the processes from step S12 to step S18, the index calculation device 20 grasps the drive voltage V when the evaluation index Qe becomes the index threshold value Qm or more (drive correction). In addition to the drive correction, or instead of the drive correction, the index calculation device 20 deteriorates the dimming sheet 10 based on the evaluation index Qe not reaching the index threshold value Qm (NO in step S14). It is also possible to notify the dimming control device 30 to that effect. At this time, the index calculation device 20 or the dimming control device 30 may store the notification as a log in order to use the state history of the dimming sheet 10 for later analysis.

The index calculation device 20 notifies the dimming control device 30 of the drive voltage V when the evaluation index Qe is equal to or higher than the index threshold value Qm. The dimming control device 30 stores the notified drive voltage V as the drive voltage V satisfying the evaluation item, and uses the stored drive voltage V when driving to satisfy the evaluation item.

On the other hand, when the index calculation device 20 determines that the current drive voltage V is the upper limit voltage Vmax (YES in step S17), the index calculation device 20 notifies the external notification device that the optical state of the dimming sheet 10 is abnormal (YES in step S17). Abnormality determination: step S19).

[Test example]
Next, a test example showing that the optical state of the dimming sheet 10 can be evaluated by image analysis will be described with reference to FIGS. 5 to 27. In the following, an example of using a polymer network type liquid crystal dimming sheet driven as a normal type as the dimming sheet 10 will be shown.

First, a device for photographing the evaluation image PE will be described. 5 and 6 show Device Example 1 and Device Example 2 used for photographing Test Examples 1 to 4.
As shown in FIG. 5, the dark room 51 of the device example 1 includes a bottom portion 51B that functions as a surface light source. The bottom portion 51B of the dark room 51 is an LED table that irradiates light toward the top of the dark room 51. The bottom of the dark room 51 supports the image target 13 inside the dark room 51. The top portion 51T of the dark room 51 is equipped with a shooting unit 12 such as a smartphone so that the inside of the dark room 51 can be photographed directly above the shooting target 13. The inner surface of the dark room 51 is black so as to prevent the photographing unit 12 from being irradiated with external light and the photographing unit 12 from being reflected.

The inside of the dark room 51 includes a mounting unit 52 for mounting the dimming sheet 10 between the photographing unit 12 and the photographing target 13. The mounting portion 52 includes a flat plate-shaped partition plate for partitioning the space in which the photographing target 13 is arranged and the other spaces, and has a rectangular hole shape at a position facing the photographing target 13 in the partition plate. A through hole is formed. The mounting portion 52 further includes a rectangular tubular mounting cylinder extending from the through hole of the partition plate toward the photographing unit 12, and the dimming sheet 10 is mounted on the mounting cylinder. The photographing unit 12 photographs the subject 13 to be photographed through the dimming sheet 10, the inside of the mounting cylinder, and the through hole of the partition plate.

The light control sheet 10 is placed on the mounting portion 52 in a state of being attached to a colorless and transparent glass plate. The dimming sheet 10 has a rectangular sheet shape, has a short side of 100 mm and a long side of 112 mm. The glass plate to which the dimming sheet 10 is attached also has a rectangular plate shape, the short side is 105 mm, and the long side is 126 mm. When the drive voltage V is 0 V, the dimming sheet 10 has the lowest transmittance, and when the drive voltage V is the upper limit voltage Vmax, the dimming sheet 10 has the highest transmittance.

The inside of the dark room 51 includes target LEDs 54R and 54L that irradiate the photographing target 13 with light L. The target LEDs 54R and 54L irradiate the entire bottom portion 51B from the lower surface of the mounting portion 52 with light L inside the dark room 51. The target LEDs 54R and 54L irradiate daylight of the same color with each other. The illuminance of the object 13 to be photographed by the object LEDs 54R and 54L is 2690 lpx.

The inside of the dark room 51 includes sheet LEDs 53R and 53L that irradiate the dimming sheet 10 with light L. The sheet LEDs 53R and 53L irradiate the entire mounting portion 52 from the top 51T of the dark room 51 with light L inside the dark room 51. The seat LEDs 53R and 53L irradiate daylight of the same color with each other. The illuminance on the dimming sheet 10 by the sheet LEDs 53R and 53L is 532 lpx.

The distance H1 in the vertical direction between the photographing unit 12 and the photographing object 13 is 350 mm. The distance H2 in the vertical direction between the photographing object 13 and the mounting portion 52 is 100 mm. The distance H3 of the mounting portion 52 in the vertical direction is 100 mm. That is, the vertical distance (= H2 + H3) between the photographing target 13 and the light control sheet 10 is 200 mm, and the vertical distance between the light control sheet 10 and the photographing unit 12 is 150 mm.

As shown in FIG. 6, the dark room 51 of the device example 2 is a calibration in which the mounting tube is omitted from the mounting portion 52 in the dark room 51 of the device example 1. The mounting portion 52 of the device example 2 includes a flat plate-shaped partition plate that partitions the space in which the photographing target 13 is arranged and the space other than the space. A rectangular hole-shaped through hole is formed in the partition plate at a position facing the photographing target 13. In the mounting section 52, the dimming sheet 10 is mounted on the partition plate. The photographing unit 12 photographs the photographing object 13 through the dimming sheet 10 and the through hole of the partition plate.

In the second apparatus example, the illuminance on the dimming sheet 10 by the sheet LEDs 53R and 53L is 1632 lx. Further, the illuminance of the object 13 to be photographed by the object LEDs 54R and 54L is 2901 lx.

In the apparatus example 2, the distance H1 in the vertical direction between the photographing unit 12 and the photographing object 13 is 350 mm, which is the same as in the apparatus example 1. On the other hand, the vertical distance (= H2) between the photographing target 13 and the light control sheet 10 is 100 mm, and the vertical distance between the light control sheet 10 and the photographing unit 12 is 200 mm.

7 (a) to 9 (a) are evaluation image examples 1-1, 1 when a test chart is taken using the apparatus example 1 and the drive voltage V is changed to 30 V, 20 V, 0 V at this time. -2, 1-3 are shown.
7 (b) to 9 (b) are evaluation image examples 2-1 and 2 when a test chart is taken using the apparatus example 2 and the drive voltage V is changed to 30 V, 20 V, 0 V at this time. -2, 2-3 are shown. The evaluation image examples 1-1, 1-2, 1-3 and the evaluation image examples 2-1, 2-2, 2-3 are all images obtained by the following processing. That is, after trimming the portion of the image taken by the photographing unit 12 through the dimming sheet 10 so that the vertical / horizontal size is 600 cps × 600 cps, the trimmed image is trimmed to have a vertical / horizontal size of 256 cps × 256 cps. It is compressed so that it becomes.

As shown in FIGS. 7 (a) to 9 (a), in the evaluation image examples 1-1, 1-2, 1-3, the opacity due to the dimming sheet 10 increases as the drive voltage V is lowered. Show that. Also in FIGS. 7 (b) to 9 (b), in the evaluation image examples 2-1, 2, 2 and 2-3, the opacity due to the dimming sheet 10 increases as the drive voltage V is lowered. show. Then, from the comparison between the evaluation image examples 1-1, 1-2, 1-3 and the evaluation image examples 2-1, 2-2, 2-3, the higher the illuminance on the dimming sheet 10, the higher the evaluation image. It can be said that the whiteness in PE increases.

In FIGS. 10 to 12, (a) shows an FFT-transformed image LP which is a two-dimensional Fourier-transformed image of evaluation image examples 1-1, 1-2, 1-3, and (b) is an evaluation image example 2-. The FFT converted image LP of 1, 2-2, 2-3 is shown.

13 (a) to 15 (a) show another evaluation image example 3-1 when a person image is taken using the device example 1 and the drive voltage V is changed to 30 V, 20 V, 0 V at this time. , 3-2, 3-3.
13 (b) to 15 (b) show another evaluation image example 4-1 when a person image is taken using the device example 2 and the drive voltage V is changed to 30 V, 20 V, 0 V at this time. , 4-2, 4-3. The evaluation image examples 3-1, 3-2, 3-3 and the evaluation image examples 4-1, 4-2, 4-3 are all images obtained by the following processing. That is, after trimming the portion of the image taken by the photographing unit 12 through the dimming sheet 10 so that the vertical / horizontal size is 600 cps × 600 cps, the trimmed image is trimmed to have a vertical / horizontal size of 256 cps × 256 cps. It is compressed so that it becomes.

As shown in FIGS. 13 (a) to 15 (a), in the evaluation image examples 3-1, 3-2, 3-3, the opacity due to the dimming sheet 10 becomes smaller as the drive voltage V is lowered. Show that it will increase. Also in FIGS. 13 (b) to 15 (b), in the evaluation image examples 4-1, 4-2, 4-3, the opacity due to the dimming sheet 10 increases as the drive voltage V is lowered. show. Then, from the comparison between the evaluation image example 3-1, 3-2, 3-3 and the evaluation image example 4-1, 4-2, 4-3, even in the other evaluation image PE, the dimming sheet 10 is used. It can be said that the whiteness in the evaluation image PE increases as the illuminance of the image increases.
16 to 18, (a) shows the FFT converted image LP of the evaluation image example 3-1, 3-2, 3-3, and (b) shows the evaluation image example 4-1, 4-2, 4-. The FFT converted image LP of 3 is shown.

[Test Example 1]
FIG. 19 is a graph showing the dependence of the drive voltage V in the radial distribution p (r), and is a test example 1 obtained from the FFT conversion image of the evaluation image example 1-1, 1-2, 1-3. The radial distribution p (r) of is shown. In FIG. 19, the value obtained by normalizing the radial distribution p (r) with the maximum value p (r) max is shown for each drive voltage V. FIG. 20 is a graph showing the dependence of the drive voltage V in the angular distribution q (θ), and is the graph of Test Example 1 obtained from the FFT conversion image of the evaluation image example 1-1, 1-2, 1-3. The angular distribution q (θ) is shown. Also in FIG. 20, the value obtained by normalizing the angular distribution q (r) with the maximum value q (θ) max is shown for each drive voltage V.

For each FFT (i) grayscale, processing according to the ITU-R Rec BT.601 standard is used, and (ii) two-dimensional fast Fourier transform and (ii) two-dimensional fast Fourier transform are used using the image data after grayscale. iii) Frequency analysis was performed in order.

As shown in FIG. 19, in the frequency analysis result of Test Example 1, a relatively large radial distribution p (r) exists in a relatively small range where the distance r is 16 or less. On the other hand, in a relatively large range where the distance r exceeds 16, there is a relatively small radial distribution p (r).

The radial distribution p (r) is lowered in the order in which the drive voltage V is lowered to 30 V, 20 V, and 0 V. The degree to which the radial distribution p (r) is lowered is small in the range where the distance r is relatively small, and large in the range where the distance r is relatively large. As described above, in the range where the distance r is relatively small, that is, in the range where the spatial frequency in the real space is relatively low, the radial distribution p (r) is unlikely to change due to the difference in the drive voltage V. On the other hand, in the range where the distance r is relatively large, that is, in the range where the spatial frequency in the real space is relatively high, the radial distribution p (r) changes greatly depending on the difference in the drive voltage V.

The range where the spatial frequency in the real space is high is a fine shade in the image after grayscale, and indicates, for example, the surface shape of the object or the outline of the object. On the other hand, the range where the spatial frequency in the real space is low is a coarse shade in the grayscaled image, and indicates, for example, the existence or nonexistence of an object. As described above, the difference in the spatial frequency in the real space appears as the difference in the radial distribution p (r) due to the difference in the drive voltage V. For example, the difference between fine shading such as the surface shape and contour of an object and coarse shading such as the presence or absence of an object appears as a difference in radial distribution p (r) due to a difference in driving voltage V.

As a result, for example, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the surface shape and contour of the object are not visually recognized, or the surface shape and contour of the object. It can be said that whether or not the optical state is such that such things are visually recognized appears as a difference in the radial distribution p (r). For example, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state such that the existence or nonexistence of an object is not visually recognized, or the optical state is such that the existence or nonexistence of an object is visually recognized. It can be said that whether or not the state is present appears as a difference in the radial distribution p (r).

In other words, the drive voltage V that makes the dimming sheet 10 opaque to the extent that the surface shape and contour of the object are not visually recognized, or the dimming sheet 10 is transparent to the extent that the surface shape and contour of the object are visually recognized. It can be said that the drive voltage V to be driven can be specified based on the fact that the sum of the radial distribution p (r) is a predetermined value. Further, the drive voltage V that makes the dimming sheet 10 opaque to the extent that the presence or absence of an object is not visually recognized, or the drive voltage V that makes the dimming sheet 10 transparent to the extent that the presence or absence of an object is visually recognized is dynamic. It can be said that it can be specified based on the fact that the sum of the radial distribution p (r) is a predetermined value.

As shown in FIG. 20, in the frequency analysis result of Test Example 1, a relatively large angular distribution q (θ) at each drive voltage V exists in the range where the angle θ is 60 ° or more and 120 ° or less. .. The integrated value in the angular distribution q (θ) in which the angle θ is 60 ° or more and 120 ° or less is lowered in the order of lowering the drive voltage V to 30V, 20V, and 0V. As described above, in the range where the angle θ is 60 ° or more and 120 ° or less, the angular distribution q (θ) greatly changes due to the difference in the drive voltage V.

As a result, it can be said that whether or not the optical state has a predetermined opacity appears as a difference in the integrated values in the angular distribution q (θ). Further, it can be said that the drive voltage V that puts the dimming sheet 10 in an optical state having a predetermined opacity can be specified based on the sum of the angular distribution q (θ) being a predetermined value.

[Test Example 3]
FIG. 21 is a graph showing the dependence of the drive voltage V in the radial distribution p (r), and is an FFT converted image of the evaluation image example 3-1, 3-2, 3-3 as in the test example 1. The radial distribution p (r) of Test Example 3 obtained from the above is shown. FIG. 22 is a graph showing the dependence of the drive voltage V in the angular distribution q (θ), and is from the FFT conversion image of the evaluation image example 3-1, 3-2, 3-3 as in the test example 1. The angular distribution q (θ) of the obtained Test Example 3 is shown.

As shown in FIG. 21, also in Test Example 3, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the surface shape, contour, etc. of the object are not visually recognized. Can be said to appear as a difference in the radial distribution p (r). Further, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the existence or nonexistence of an object is not visually recognized appears as a difference in the radial distribution p (r). It can be said that.

As shown in FIG. 22, even in Test Example 3, it can be said that whether or not the optical state has a predetermined opacity appears as a difference in the integrated value in the angular distribution q (θ). Further, it can be said that the drive voltage V that puts the dimming sheet 10 in an optical state having a predetermined opacity can be specified based on the sum of the angular distribution q (θ) being a predetermined value.

[Test Example 2]
FIG. 23 is a graph showing the dependence of the drive voltage V in the radial distribution p (r), and is an FFT conversion image of the evaluation image examples 2-1, 2, 2 and 2-3 as in the test example 1. The radial distribution p (r) of Test Example 2 obtained from the above is shown. FIG. 24 is a graph showing the dependence of the drive voltage V in the angular distribution q (θ), from the FFT conversion images of the evaluation image examples 2-1, 2, 2-3, as in the test example 1. The angular distribution q (θ) of the obtained Test Example 2 is shown.

As shown in FIG. 23, also in Test Example 2, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the surface shape, contour, etc. of the object are not visually recognized. Can be said to appear as a difference in the radial distribution p (r). Further, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the existence or nonexistence of an object is not visually recognized appears as a difference in the radial distribution p (r). It can be said that.

As shown in FIG. 24, even in Test Example 2, it can be said that whether or not the optical state has a predetermined opacity appears as a difference in the integrated value in the angular distribution q (θ). Further, it can be said that the drive voltage V that puts the dimming sheet 10 in an optical state having a predetermined opacity can be specified based on the sum of the angular distribution q (θ) being a predetermined value.

[Test Example 4]
FIG. 25 is a graph showing the dependence of the drive voltage V in the radial distribution p (r), and is an FFT converted image of the evaluation image example 4-1, 4-2, 4-3 as in the test example 1. The radial distribution p (r) of Test Example 4 obtained from the above is shown. FIG. 26 is a graph showing the dependence of the drive voltage V in the angular distribution q (θ), and is from the FFT conversion image of the evaluation image example 4-1, 4-2, 4-3 as in the test example 1. The angular distribution q (θ) of the obtained Test Example 4 is shown.

As shown in FIG. 25, also in Test Example 4, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the surface shape, contour, etc. of the object are not visually recognized. Can be said to appear as a difference in the radial distribution p (r). Further, whether or not the dimming sheet 10 to which a predetermined drive voltage V is applied is in an optical state to the extent that the existence or nonexistence of an object is not visually recognized appears as a difference in the radial distribution p (r). It can be said that.

As shown in FIG. 26, even in Test Example 4, it can be said that whether or not the optical state has a predetermined opacity appears as a difference in the integrated value in the angular distribution q (θ). Further, it can be said that the drive voltage V that puts the dimming sheet 10 in an optical state having a predetermined opacity can be specified based on the sum of the angular distribution q (θ) being a predetermined value.

[Haze and evaluation index Qe]
FIG. 27 shows the dependence of the evaluation index Qe obtained from the radial distribution p (r) of Test Example 1 to Test Example 4 on the drive voltage V, and the dependence of the haze of the dimming sheet 10 on the drive voltage V. Is shown. Note that FIG. 27 shows the evaluation index Qe standardized by the reference value obtained by dividing the sample integration value SumV calculated at each drive voltage V by the reference value. At this time, the sample integrated value SumV when the drive voltage V is 30 V, that is, the sample integrated value SumV when the dimming sheet 10 has the highest light transmittance was used as a reference value. Further, in FIG. 27, the haze of the dimming sheet 10 shows a value standardized with 100% as 1.

As shown in FIG. 27, the haze of the dimming sheet 10 shows a substantially constant value close to 100% in the range where the drive voltage V is 0 V or more and 10 V or less. Further, in the range where the drive voltage V is 0 V or more and 10 V or less, each evaluation index Qe from Test Example 1 to Test Example 4 also shows a substantially constant value of 0.2 or less.

In the range where the drive voltage V is 15 V or more and 20 V or less, the haze of the dimming sheet 10 drops sharply from 65% to 30%. Further, in the range where the drive voltage V is 15 V or more and 20 V or less, each evaluation index Qe from Test Example 1 to Test Example 4 sharply increases from 0.2 to 0.8.

In the range where the drive voltage V is 25 V or more and 30 V or less, the haze of the dimming sheet 10 gradually decreases from 20% to 10%. Further, in the range where the drive voltage V is 25 V or more and 30 V or less, each evaluation index Qe from Test Example 1 to Test Example 4 gradually increases from 0.8 to 1.0.

In this way, the evaluation index Qe changes according to the change in haze in the dimming sheet 10 in the entire range where the drive voltage V is 0 V or more and 30 V or less. Further, the evaluation index Qe is maintained following the maintenance of the haze in the dimming sheet 10. As a result, it can be said that the evaluation index Qe obtained by using the image analysis of the evaluation image PE represents the optical state of the dimming sheet 10.

As described above, according to the above embodiment, the effects listed below can be obtained.
(1) The degree of blurring of the evaluation image PE taken by the light control sheet 10 by the light control sheet 10 is calculated by using the image analysis of the evaluation image PE. Since the shooting of the evaluation image PE is realized not by a dedicated optical device such as an integrating sphere but by a general-purpose shooting device such as a smartphone or a tablet terminal, the general-purpose image PE is used when evaluating the optical state of the dimming sheet 10. It is possible to enhance the sex.

(2) The degree of blurring by the light control sheet 10 tends to increase as the haze of the light control sheet 10 increases, and tends to change in synchronization with the haze change of the light control sheet 10. Therefore, with the above configuration for calculating the degree of blurring of the evaluation image PE, it is possible to diversify the evaluation index representing the optical state of the light control sheet 10.

(3) Whether the dimming sheet 10 is normal by using the index threshold value Qm, which is the evaluation index Qe when the optical state of the dimming sheet 10 is normal, and the evaluation index Qe calculated by the index calculation device 20. Whether or not it is determined. Therefore, it is possible to determine whether or not the dimming sheet 10 is normal by using the new evaluation index Qe.

(4) Evaluation image The sharpness of repetition in a short cycle such as the surface shape and contour of an object in PE is detected as the magnitude of a frequency component in a high frequency band. On the other hand, the sharpness of repetition in a long period such as the presence or absence of an object in the evaluation image PE is detected as the magnitude of the frequency component in the low frequency band. In this regard, if the evaluation index Qe is calculated through the frequency analysis of the evaluation image PE, the optical state of the dimming sheet 10 can be represented in detail by the evaluation index Qe.

(5) A sample group for calculating the evaluation index Qe is extracted based on the sampling conditions, and the above conditions 1 and 2 are set as the sampling conditions. That is, a predetermined frequency band in which the sample threshold value rx is larger than the extraction reference distance rs is extracted as a sample group for calculating the evaluation index Qe. This makes it possible to provide, for example, an evaluation index Qe specialized for the sharpness of repetition in a short cycle.

If a predetermined frequency band in which the sample threshold value rx is equal to or less than the extraction reference distance rs is included in the sample group for calculating the evaluation index Qe, for example, an evaluation index that takes into account the sharpness of repetition in a long period. It is also possible to provide Qe.

(6) Since the evaluation index Qe is calculated from the sample integrated value SumV in the radial distribution p (r), the spatial frequency component appears in the one-dimensional direction or the spatial frequency component appears in the two-dimensional direction. That is, it is possible to calculate the evaluation index Qe having a size according to the optical state of the dimming sheet 10. That is, the restrictions related to spatial anisotropy are alleviated in the photographing target 13. Therefore, it is possible to expand the situations in which the index calculation device 20 is applied in various fields, and to increase the versatility of the index calculation device 20.

(7) The radial distribution p (r) of a relatively short distance r tends to indicate only the existence or nonexistence of an object through the dimming sheet 10, and tends to indicate the maximum energy. In this regard, as indicated by the above conditions 1 and 2, the evaluation index Qe is calculated with the band higher than the maximum value p (r) max as the sample group, so that the visibility is not rough as to whether or not the object exists. It is also possible to evaluate whether or not the dimming sheet 10 provides fine visibility such as the contour of an object.

(8) The sample integrated value SumV when the drive voltage V is 0 V is standardized by the sample integrated value SumV when the drive voltage V is 30 V, and the standard value is calculated as the evaluation index Qe. Therefore, it is possible to evaluate the optical state of the dimming sheet 10 when the driving voltage V is not applied, with reference to the optical state of the dimming sheet 10 when the driving voltage V is applied. Therefore, it is possible to evaluate the optical state of the dimming sheet 10 including whether or not the dimming control device 30 is normal.

The above embodiment can also be modified and implemented as follows.
The evaluation index Qe is not limited to the value in which the sample integrated value SumV is standardized by the reference value, and may be, for example, the sample integrated value SumV itself. If the configuration is such that the evaluation index Qe when the dimming sheet 10 is transparent is calculated and the other evaluation index Qe is standardized using the evaluation index Qe as a reference value, the difference in lighting at the time of shooting and the shooting target The error due to the difference of 13 can be reduced in the evaluation index Qe. Then, it is possible to calculate the degree of blurring by the dimming sheet 10 with high accuracy.

The index calculation device 20 may include an output unit in which the communication unit 24 is omitted and the evaluation index Qe calculated by the image analysis unit 22 is displayed or output by voice. At this time, in the processing using the evaluation index Qe such as normality determination, abnormality determination, and drive correction from step S14 to step S18, the dimming control device 30 is configured so that the evaluation index Qe is input from the operation unit or the like. It may be done with other devices such as.

p (r) ... Radial radial distribution PE ... Evaluation image Qe ... Evaluation index 10 ... Dimming sheet 12 ... Shooting unit 13 ... Shooting target 20 ... Index calculation device 22 ... Image analysis unit 23A ... Evaluation index calculation program 30 ... Dimming Control device

Claims (10)

It is an evaluation index calculation device that calculates an evaluation index that represents the optical state of the dimming sheet.
An image of a subject in a predetermined environment taken through the dimming sheet is defined as an evaluation image, and the degree of blurring of the evaluation image by the dimming sheet is used as the evaluation index by using image analysis of the evaluation image. An evaluation index calculation device provided with a calculation unit for calculation. A determination unit for determining whether or not the light control sheet is normal by using the evaluation index when the optical state of the light control sheet is normal and the evaluation index calculated by the evaluation index calculation device. The evaluation index calculation device according to claim 1. The evaluation index calculation device according to claim 1, wherein the image analysis is a frequency analysis of the evaluation image. The evaluation index calculation device according to claim 3, wherein the degree of blurring is the magnitude of a frequency component in a predetermined frequency band. The evaluation index calculation device according to claim 4, wherein the magnitude of the frequency component is the magnitude of the radial distribution obtained from the Fourier transform of the image. The evaluation index calculation device according to claim 4 or 5, wherein the predetermined frequency band is a band higher than the frequency at which the energy shows the maximum in the radial distribution obtained from the Fourier transform of the image. The magnitude of the frequency component when the optical state of the dimming sheet is normal is stored as a normal range, and when the magnitude of the frequency component is within the normal range, the optical state of the dimming sheet is normal. The evaluation index calculation device according to any one of claims 3 to 6, which is determined to be present. A value normalized by the degree of blurring of the evaluation image when the driving voltage is not applied to the dimming sheet is calculated by the degree of blurring of the evaluation image when the driving voltage is applied to the dimming sheet. The evaluation index calculation device according to any one of claims 1 to 7. It is an evaluation index calculation method that calculates an evaluation index that represents the optical state of the dimming sheet.
An image of a subject in a predetermined environment taken through the dimming sheet is defined as an evaluation image, and the degree of blurring of the evaluation image by the dimming sheet is used as the evaluation index by using image analysis of the evaluation image. Evaluation index calculation method including calculation. For an evaluation index calculation device that calculates an evaluation index that represents the optical state of the dimming sheet,
An image of a subject in a predetermined environment taken through the dimming sheet is defined as an evaluation image, and the degree of blurring of the evaluation image by the dimming sheet is used as the evaluation index by using image analysis of the evaluation image. Evaluation index calculation program to be calculated.

Images