JP7076760B2 - Color judgment device and color judgment method - Google Patents

Description

The present invention relates to a color determination device and a color determination method for determining a color / texture by comparing a color standard body with an inspection target.

According to the color fidelity environment correction device or the color fidelity environment correction method of Patent Document 1, it is possible to provide a simulation by replacing the color of a product in the sale of a product such as a car under conditions where there is a difference in environment, which is convenient for manufacturing and sales. In addition to being able to provide the above, the color of the purchased product will be similar to the place of product selection from the human eye, so there is an effect of reducing the discomfort of the color.

The wearable coloring evaluation device and the wearable coloring evaluation method of Patent Document 2 are, in an RGB color camera used in a general wearable terminal and an industrial ITV, the color data in the RGB color range is the colorimetric data of the XYZ color function. Accurate color conversion from RGB color space to XYZ color measurement space in order to have a correction function that can be used. For example, a colorimeter with RGB color that can perform conversion with a color difference (for example, ΔE <1.6). Provide a side system.

WO2015 / 141233A1 Publication Japanese Unexamined Patent Publication No. 2016-6408

Currently, in the transportation industry, the electrical industry, etc., a method of visually comparing using a standard plate (standard color sample determined by the design department) in order to match the color determined by the design department in the product color inspection in the manufacturing department. Is adopted.

However, the color environment correction device of Patent Document 1, the wearable coloring evaluation device of Patent Document 2, and the wearable coloring evaluation method have not yet been sufficiently compatible with the standard plate.

For example, in the automobile industry, there are two major issues: "too many standard plates" due to the diversification of vehicle types and color variations, and "differences in the appearance of standard plates due to ambient light". It is not used effectively and relies on the experience and intuition of experts. If the colors of the vehicle do not match in this method, it is necessary to correct the color of the vehicle in the correction process, which may cause a lot of quality loss (see FIG. 14). In a manufacturing factory or the like, products of many colors flow into one line in an actual production line, and in this situation, comparison with a standard plate is visually performed, so that confirmation work is difficult.

For this reason, it is expected to develop a tool that allows general workers to acquire highly skilled color / texture recognition ability. Under such circumstances, there is a need to share the color data of the standard plate between the design department and the manufacturing department and use it for color inspection.

In the manufacturing process in the automobile industry, etc., "there are too many standard plates", so it takes too much time to visually correct the color based on the standard plate at the assembly stage. It is carried out. If color correction is possible at the assembly stage, quality loss can be prevented and the subsequent correction process can be omitted, so there is a high need for it. At the same time, it can be expected that similar needs will be demanded in the home appliance industry and the construction industry with the spread of information processing technology.

The subject of the present invention is a two-dimensional colorimeter using a wearable terminal as a tool that can follow a technique comparable to that of a skilled person in the process of performing a color inspection by comparing with a standard plate with the eyes of a skilled person. The purpose is to provide an electronic replacement tool for color standard plates by incorporating measurement technology and information processing technology.

The subject of the present invention is to display the color information of a product electronically in an ambient light on a wearable terminal by using information processing technology, and visually check it while performing assembly work, etc., and a highly skilled person. It is to provide a tool to give general workers the ability to recognize the color and texture of. In fields where quality control has been advanced with the eyes of experts (painting process in automobiles, home appliances, adjustment process, etc.), technical transfer has become a major issue, and in such processes, it is comparable to experts. The purpose is to contribute to the manufacturing industry toward the sophistication of technology in the information processing field by developing and utilizing wearable terminals as a tool that can follow the technology to be processed.

An object of the present invention is that in the assembly process during the manufacturing process, an accurate standard plate is projected on the display unit of the wearable terminal, and the standard plate and the inspected product can be viewed at the same time to perform color correction at the assembly stage. It is possible to improve the accuracy of color determination, eliminate the need for a subsequent correction process, and enable color inspection using a standard plate not only for skilled workers but also for general workers.

In view of the above problems, the color determination device of the present invention has a display unit for displaying an image, a wearable terminal on the remote side that receives a signal and displays the image on the display unit, and a color temperature under remote illumination light. A two-dimensional color with a color temperature measurement sensor on the remote side and three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) linearly converted equivalent to the CIE XYZ color matching function. The wearable terminal is equipped with a local IOT server that stores a plurality of 3-band visual sensitivity images S1i, S2i, and S3i that image a plurality of color standard bodies under a standard light source. A camera is provided, and the arithmetic unit obtains the three-band visual sensitivity images S1i, S2i, and S3i stored in the IOT server based on the integrated value of the measured values of the color temperature measuring sensor. It has an illumination color conversion correction unit that adjusts and converts to an XYZ correction value with color temperature correction, and the IOT server stores the tag information of the product to be color-determined and corresponds to the tag information. The wearable terminal receives the XYZ correction value of the color standard body from the illumination color conversion correction unit, displays a correction image corresponding to the XYZ correction value on the display unit, and uses a small camera of the wearable terminal. The product to be the target of the color determination is imaged, the calculation device measures the color distribution of the captured product image, the corrected image and the product image are compared, and the color and the texture spread index are separated and calculated. Then, the comparison result is displayed on the display unit, and the chromaticity diagram obtained by superimposing the chromaticity histogram distribution or the color space histogram of the product image and the corrected image on the chromaticity diagram is displayed on the display unit. It features a color determination device that can be used . IOT is an abbreviation for internet of things.

The two-dimensional colorimeter adopts a multi-point simultaneous measurement method using an image sensor, which enables evaluation of color and texture, and can be used in the automobile industry, home appliance industry, cosmetics industry, printing industry, building materials industry, and the like.

The inspection area of the coordinates corresponding to the color space of the XYZ color system based on the three-band visual sensitivity images S1i, S2i, and S3i is divided by a grid, and the number of pixels of the inspection object and the color standard belonging to each grid are integrated. By doing so, a color space histogram distribution of the XYZ color system is created, the centers of the two color space histogram distributions of the inspection object and the color standard are specified, and the center of one of the color space histogram distributions is set to the other. It is preferable to calculate the texture spread index indicating the difference in the spread of the color space histogram distribution by shifting the color space histogram distribution so as to be close to the center.

By imaging the product with a two-dimensional colorimeter on the remote side, shifting the color and texture of the product and the standard plate to separate them, and displaying this on the remote PC for comparison, it is usually a wearable terminal. However, if the color judgment is delicate with the human eye using a wearable terminal, an advanced version that allows the comparison of the texture with the standard plate to be referred to is also possible. It is preferable from the viewpoint of shortening the working time.

The color determination method of the present invention is a standard light source using a two-dimensional colorimeter having three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) linearly converted equivalent to the CIE XYZ color matching function. Below, a storage step for storing a plurality of 3-band visual sensitivity images S1i, S2i, S3i in which a plurality of color standard bodies are imaged, and a color temperature measurement step for measuring a color temperature with a color temperature measurement sensor under remote illumination light. Then, the arithmetic unit of the wearable terminal adjusts the gain of the two-dimensional colorimeter based on the integrated value of the measured values of the color temperature measurement sensor for the three-band visual sensitivity images S1i, S2i, S3i, and colors. The illumination color conversion correction step for converting to a temperature-corrected XYZ correction value, and the IOT server stores the tag information of the product to be color-determined, and the XYZ correction value of the color standard corresponding to the tag information is stored. The wearable terminal receives the image, displays the corrected image corresponding to the XYZ correction value on the display unit of the wearable terminal, and images the product to be color-determined by the small camera of the wearable terminal. The arithmetic unit measures the color distribution of the captured product image, compares the corrected image with the product image, calculates the color and the texture spread index separately, and displays the comparison result on the display unit. It is characterized by comprising a display step of displaying a chromaticity diagram in which the product image and the chromaticity histogram distribution or the color space histogram of each of the corrected images are superimposed on the chromaticity diagram on the display unit. ..

The coordinate inspection area corresponding to the color space of the XYZ color system based on the three-band visual sensitivity images S1i, S2i, and S3i is divided by a grid, and the number of pixels of the inspection object and the color standard belonging to each grid is determined. By integrating, the step of creating the color space histogram distribution and the center of the two color space histogram distributions of the inspection object and the color standard are specified, and the center of one of the color space histogram distributions is set to the other color. It is preferable to include a step of shifting the color space histogram distribution so as to be close to the center and calculating a texture spread index indicating the difference in the spread of the color space histogram distribution.

Examples of the wearable terminal include a wearable glass having a display unit on glasses, a wearable device having a head-mounted display, and the like.

As the standard light source, for example, a xenon lamp, which is an illumination close to the artificial sun, can be exemplified as a light source.

It is preferable that the IOT server stores the tag information of the product and includes a step of displaying an image corresponding to the XYZ value of the product or the standard body corresponding to the tag information on the display unit.

In the color space, the size that can be specified by three independent quantities is expressed by the distance from the origin in the three axial directions, and the color can be specified as the coordinates of one point in the space.

By simply attaching the wearable terminal to a person and displaying the standard color on the wearable terminal and comparing it with the actual product, the skill level of the color judgment work by the standard body is dramatically improved without obstructing the flow of other work. It will be possible to improve. Since the present invention has color environment correction as compared with the conventional color inspection, it has a high quality advantage and incorporates measurement technology and information processing with a two-dimensional colorimeter using a wearable terminal. However, by making it a wearable terminal, it is possible to use it as an alternative tool to the standard board, so there is an advantage that even a worker with low skill in color judgment can immediately perform an advanced inspection on color judgment.

It is possible to reduce costs by improving the efficiency of the color matching process and reducing waste of color correction. In addition, the introduction cost is low, and it is highly accurate and stable.

It is a block diagram of the color determination apparatus 1 of Embodiment 1 of this invention. It is a block diagram in the case of performing the color correction by the micro optical system of the color determination apparatus 1 of Embodiment 1 of this invention. Similarly, it is a perspective view of the modified form of the head-mounted display type wearable terminal 3 of the color determination device 1. It is a block diagram of the 2D colorimeter of the color determination apparatus 1 of Embodiment 1 of this invention. It is a specific example of the method of acquiring image information according to three spectral sensitivities in Embodiment 1 of this invention. (A) is an explanatory diagram when a dichroic mirror is used. (B) is an explanatory diagram when a filter turret is used. (C) is an explanatory diagram when the optical filter is microscopically attached to the image pickup device 53. It is a function showing the spectral sensitivity of the two-dimensional colorimeter 2 which is the XYZ color system camera in the first embodiment of the present invention. It is a block diagram of the IOT server of the color determination apparatus of Embodiment 1 of this invention. It is a flowchart in 2D colorimeter of Embodiment 1 of this invention. It is a flowchart in the IOT server of Embodiment 1 of this invention. It is a flowchart of the process of the illumination color conversion correction part of Embodiment 1 of this invention. It is explanatory drawing of the process of the illumination color conversion correction part of Embodiment 1 of this invention. It is explanatory drawing of the process of the illumination color conversion correction part of Embodiment 1 of this invention. It is explanatory drawing which shows the application to the painting process of the color determination apparatus of Embodiment 1 of this invention. It is explanatory drawing of the conventional painting process. It is a perspective view of the wearable terminal of Embodiment 2 of this invention. It is a flowchart in the controller of the wearable terminal of Embodiment 2 of this invention. It is a sub-flow chart in the controller of the wearable terminal of Embodiment 2 of this invention. It is explanatory drawing which shows the shift process in the xy coordinate space in the controller of Embodiment 2 of this invention. (A) is an explanatory diagram showing an inspection region T in the controller of the second embodiment of the invention, (b) is an xy chromaticity diagram showing a chromaticity region K on a chromaticity diagram corresponding to the inspection region T, and (c) is a grid. An explanatory diagram of the chromaticity region K partitioned by G, (d) is a schematic diagram showing the state of overlapping chromaticities on the xy two-dimensional chromaticity diagram, (e) is an explanatory diagram showing a minimum distribution, (f). Is an explanatory diagram showing an example of the xy chromaticity histogram distribution (the numerical values of the squares are the count values of the pixels). (A) is an explanatory diagram showing the metallic degree of bumper B, (b) is an xy chromaticity histogram distribution diagram, and (c) is a three-dimensional image diagram of an xy chromaticity histogram distribution. It is a block diagram which shows the structure of the color determination apparatus of Embodiment 3 of this invention. It is a flowchart (XYZ color space distribution) in the calculation part of the color determination apparatus of Embodiment 3 of this invention. It is a flowchart (Lab color space distribution) in the calculation part of the color determination apparatus of Embodiment 3 of this invention. It is explanatory drawing which shows the shift process in the XYZ coordinate space of the color determination apparatus of Embodiment 3 of this invention. It is a block diagram which shows the structure of the color determination apparatus of Embodiment 4 of this invention. It is explanatory drawing which shows the shift process in the Lab coordinate space of the color determination apparatus of Embodiment 4 of this invention. It is explanatory drawing which shows the processing by the color determination apparatus of Embodiment 5 of this invention.

The color determination device 1 according to the first embodiment of the present invention will be described with reference to the drawings. FIG. 1 shows an example in which the color determination device 1 is applied to the assembly process of the bumpers B ₁ , B ₂ , B ₃ ... To the vehicles P ₁ , P ₂ , P ₃ ... Of the production line. .. If the suffixes of the car bumper codes are the same, they are the same color. The car P ₁ and the bumper B ₁ are red, the car P ₂ and the bumper B ₂ are blue, and the car P ₃ and the bumper B ₃ are yellow.

The color determination device 1 has an RGB display unit 2 that displays an RGB image, a wearable terminal 3 on the remote side that receives an RGB signal and displays the RGB image on the RGB display unit 2, and a color under remote illumination light. A color temperature measuring sensor 4 on the remote side that measures the temperature, and two having three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) linearly converted equivalent to the CIE XYZ color matching function. Local side that stores a plurality of 3-band visual sensitivity images S1i, S2i, S3i obtained by color-imaging a plurality of vehicles N1 or a standard plate N2 of a product under a standard light source (for example, the standard light is D65) by the dimensional color meter 5. IOT server 6 and the 3-band visual sensitivity images S1i, S2i, S3i stored in the IOT server 6 are converted into corrected images with color temperature correction based on the measured values of the color temperature measurement sensor 4. The correction unit 7 and the wearable terminal 3 receive a signal from the illumination color conversion correction unit 7 and display the correction image on the RGB display unit 2. The connection method of each part can be selected regardless of whether it is wired or wireless.

The factory control computer (hereinafter abbreviated as control PC) 8 reads and stores the information of the IC tag T of the product flowing through the factory production line, sends this information to the IOT server 6, and the IC tag T. The three-band visual sensitivity images S1i, S2i, and S3i corresponding to the above are read out, the color temperature is corrected by the illumination color conversion correction unit 7, the XYZ value is obtained, converted into an RGB value, and this is transmitted to the wearable terminal 3. Then, it is displayed on the RGB display unit 2. The image displayed on the RGB display unit 2 is updated every time the color / texture of the vehicles P ₁ , P ₂ , P ₃ ... In the assembly process is changed.

It is also possible to correct the color and texture displayed on the RGB display unit 2. For example, as shown in FIG. 2, the micro optical system of the RGB display unit 2 which is an RGB color display incorporated in the wearable terminal 3 displays candidate RGB color / texture data and signals in each RGB channel. The linearity characteristics and color development / texture characteristics of the above are measured, and the correction terms such as crosstalk of the color / texture data in this color mixture are calculated. At the same time, the XYZ color space measurement is performed by the micro optical system reading spectroscope (small high-precision spectroscope) 9, and the high-precision XYZ correction data is acquired. The system control unit 10 corrects the XYZ data from the compact high-precision spectroscope 9 by the color / texture correction unit 11, and stores the correction data in the storage unit 12. A signal is output from the system control unit 10 to the signal pattern generator 13, and this signal is transmitted to the RGB display unit 2.

When the XYZ image data measured by the two-dimensional colorimeter 5 is displayed on the display unit 2 of the wearable terminal 3, the difference from the display on the high color rendering monitor 15 is verified, and these displays are corrected. .. Consider that this correction method is a micro optical system. Since this high color rendering correction method is arbitrary, it may or may not be added to the ambient light correction method.

For example, an object shot in a shooting environment in a studio in the design department and whose color is confirmed by a high color performance monitor 15 is measured in color temperature by a color temperature measuring sensor 4 with illumination light assuming a remote environment, and based on that information. The wearable terminal 3 may display an image obtained by performing color conversion in a remote environment from the correction data of the wearable terminal 3.

Next, each part of the color determination device 1 described above will be described in detail.

The wearable terminal 3 is provided with an RGB display unit 2 on eyeglasses, and is also provided with a control unit, an input / output interface, and the like.

As a modification of the wearable terminal 3, a head-mounted display type wearable terminal 3'is also possible. By projecting the image from the light source into the eyes, this wearable terminal 3'appears that the image floats in front of the eyes, and by projecting the image at a close distance to the eyeball, the posture, movement, and space are not restricted. It can receive video data. Since it is very small unlike a normal display, it has excellent portability and power saving. Although it is small, it can be used as a virtual large display because it is very close to the eyeball, and because it is worn on the head, it has high followability that exists in the field of view. Images can be projected without wearing glasses. It can also be used over ordinary glasses or goggles. You can work while looking at the screen, and you do not have to move your eyes significantly when checking information. The wearable terminal 3'includes a headband 32 provided with an arm band 31, a control box 33 equipped with an input terminal and a battery, and an RGB display unit 2'.

The color temperature measurement sensor 4 includes a micro spectroscope, and calculates and outputs a color temperature from the measurement data of this spectroscope. As this micro spectroscope, for example, C12666MA manufactured by Hamamatsu Photonics Co., Ltd. can be exemplified. This is a fingertip-sized ultra-compact spectroscope head that combines MEMS technology and image sensor technology, with a sensitivity wavelength range of 340 to 780 nm and a wavelength resolution of 15 nm max. The object is imaged by the micro color temperature measurement sensor 4, and the spectral sensitivity characteristic, that is, the relative sensitivity (%) which is the output value of the spectrum with respect to the wavelength in the sensitivity wavelength range is obtained.

The two-dimensional colorimeter 5 is a product number RC-500 of Paparabo Co., Ltd., and as shown in FIG. 3, three spectral sensitivities (S1 (λ) and S2 (λ)) linearly converted equivalent to the CIE XYZ color matching function. , S3 (λ)), and includes an arithmetic processing unit 54 that converts the captured image into three spectral sensitivities into a tristimulus value XYZ in the CIE XYZ color system, and an image display unit 55 that displays the image. There is.

The method shown in FIG. 5A is a method using a dichroic mirror. This reflects light of a specific wavelength by the dichroic mirror 52c', and the remaining light transmitted is reflected and dispersed by another dichroic mirror 52a'of another specific wavelength, and the image pickup elements 53a, 53b , 53c are read out in parallel. Here, the dichroic mirror 52a'corresponds to the optical filters 52a and 52b, and the dichroic mirror 52c' corresponds to the optical filter 52c. As for the light incident from the photographing lens 51, the light according to the spectral sensitivity S3 is reflected by the dichroic mirror 52c', and the remaining light is transmitted. The light reflected by the dichroic mirror 52c'is reflected by the reflecting mirror 56, and the spectral sensitivity S3 is obtained by the image pickup element 53c. On the other hand, as for the light transmitted through the dichroic mirror 52c', the light according to the spectral sensitivity S1 is reflected by the dichroic mirror 52a', and the light according to the remaining spectral sensitivity S2 is transmitted . The light transmitted through the dichroic mirror 52a'is imaged by the image pickup device 53b to obtain the spectral sensitivity S2. The light reflected by the dichroic mirror 52a'is reflected by the reflecting mirror 59, and the spectral sensitivity S1 is obtained by the image pickup element 53a . Instead of the dichroic mirror, a dichroic prism having the same characteristics may be used for splitting into three, and the image pickup elements 53a, 53b, and 53c may be adhered to the positions where the respective lights are transmitted.

The method shown in FIG. 5B is a method using the filter turret 57. Optical filters 52a, 52b, 52c are provided on the filter turret 57 having the same direction as the incident light from the photographing lens 51 on the rotation axis, and these are mechanically rotated. S1, S2, and S3 are obtained.

The method shown in FIG. 5C is a method in which the optical filters 52a, 52b, 52c are microscopically attached to the image pickup device 53. The optical filters 52a, 52b, 52c on the image pickup device 53 are provided in a Bayer array type. In this arrangement, the optical filter 52b is provided in half of the region on the image pickup device 53 divided in a grid pattern, and the optical filter 52a and the optical filter 52c are evenly arranged in the other half region. That is, the arrangement amount is optical filter 52a: optical filter 52b: optical filter 52c = 1: 2: 1. It is not particularly hindered in the first embodiment that the arrangement of the optical filters 52a, 52b, 52c is other than the Bayer arrangement. Since each of the optical filters 52a, 52b, and 52c is very fine, they are attached to the image pickup device 53 by printing. However, the present invention is not meaningful in this arrangement, but is to attach a filter having characteristics of spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)) to the image pickup device.

The spectral sensitivity of the two-dimensional colorimeter 5 satisfies the router condition, and the spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)) is obtained from the XYZ color function as shown in FIG. , It is a mountain shape that does not have a negative value and has a single peak, and is equivalently converted from the condition that the peak values of the respective spectral sensitivity curves are equal and the overlap of the spectral sensitivity curves is minimized. Specifically, the spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)) has the following characteristics.
Peak wavelength Half width 1/10 width S1 582nm 523-629nm 491-663nm
S2 543nm 506-589nm 464-632nm
S3 446nm 423-478nm 409-508nm

The peak wavelength of the spectral characteristic S1 can be treated as 580 ± 4 nm, the peak wavelength of the spectral characteristic S2 can be treated as 543 ± 3 nm, and the peak wavelength of the spectral characteristic S3 can be treated as 446 ± 7 nm.

The spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) of the two-dimensional colorimeter 5 are mountain shapes having a single peak that do not have a negative value from the CIE XYZ spectral characteristics, and their respective spectral sensitivities. Equivalent conversion is performed from the condition that the peak values of the curves are equal and the overlap of the spectral sensitivity curves is minimized. The curve of the spectral characteristic S1 has a peak wavelength of 582 nm and a half width of 523 to 629 nm. The width of 1/10 is 491 to 663 nm. The curve of the spectral characteristic S2 has a peak wavelength of 543 nm, a half width of 506 to 589 nm, and a 1/10 width of 464 to 632 nm. The curve of the spectral characteristic S3 has a peak wavelength of 446 nm, a half width of 423 to 478 nm, and a 1/10 width of 409 to 508 nm.

The three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) are obtained by using Equation 1. For details on the spectral characteristics themselves, refer to Japanese Patent Application Laid-Open No. 2005-257827.

This color fidelity technology can reproduce the degree and degree of reproduction of the original color / texture from the information about the color / texture obtained from the outside by devices such as cameras, displays, and printers. With regard to color and texture, it is possible to obtain significantly reproducibility as compared with the conventional RGB color method.

The specifications of the two-dimensional color meter 5 are, for example, an effective frequency value of about 5 million pixels, an effective area of 9.93 mm x 8.7 mm, an image size of 3.45 μm x 3.45 μm, a video output of 12 Bit, a camera interface GigE, and the number of frames (when focusing is adjusted). 3 to 7 frames / sec, shutter speed 1 / 15,600sec to 1/15sec, integrated time up to 3 seconds, S / N ratio 60dB or more, lens mount F mount, operating temperature 0 ° C to 40 ° C, operating humidity 20% to 80 %.

As shown in FIG. 4, the two-dimensional colorimeter 5 includes a photographing lens 51, three optical filters 52a, 52b, 52c arranged behind the photographing lens 51, and behind the optical filters 52a, 52b, 52c. It includes an arranged image pickup element 53 (CCD, CMOS, etc.). The three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) of the two-dimensional colorimeter 5 are determined by the product of the spectral transmittances of the optical filters 52a, 52b and 52c and the spectral sensitivities of the image pickup element 53. It is given. The arrangement relationship between the optical filters 52a, 52b, 52c and the image pickup device 53 in FIG. 4 is merely schematically shown. Specific examples of the method of acquiring image information according to the three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) are given below, but any of these can be adopted in the present embodiment. , Also, other methods can be adopted.

The two-dimensional colorimeter 5 includes an arithmetic processing unit 54, and three-band visual sensitivity images S1i, S2i, S3i (T = 6500K) are produced by spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)). The acquired, recorded, and visualized image is displayed on the image display unit 55, and is transmitted to the IOT server 6 via a communication line such as the Internet. The transmission is not limited to the communication line, and may be performed by transporting a storage medium, a wired connection, or other means.

The IOT server 6 is connected to the two-dimensional colorimeter 5, the illumination color conversion unit 7, and the control PC 8, and can communicate with the wearable terminal 3 and the color temperature measurement sensor 4 via the Internet. The IOT server 6 includes an input / output interface 61, a calculation unit 62, a storage unit 63, a display unit 64, and a bus line 65. The arithmetic unit 62 includes a CPU, a ROM, and a RAM. The storage unit 63 is a hard disk.

The illumination color conversion unit 7 is a control computer that processes a corrected image, and is provided with a CPU, a ROM, a RAM, an input / output interface, and the like. The illumination color conversion unit 7 may be built in the IOT server 6 and integrated.

The control PC 8 performs measurement control processing of the production line, and is provided with a CPU, ROM, RAM, an input / output interface, and the like.

It can be applied regardless of whether the two-dimensional colorimeter 2, the wearable terminal 3, or the IOT server 6 is located at a remote location or at the same location. Here, the case where the wearable terminal 3 is in a remote place will be described. The design department is on the local side, and the vehicle N1 is imaged by the two-dimensional colorimeter 5, and the three-band visual sensitivity images S1i, S2i, S3i (T = 6500K) (i = 1 to m) are stored in the storage unit 63 of the IOT server 6. , M is the number of pixels.) Is recorded. T = 6500K under the local environment, T = 2900K under the remote environment, and the illumination color converter 7 corrects the 3-band visual sensitivity images S1i, S2i, and S3i based on the difference in the color temperature of the ambient light. conduct.

The configuration and operation of the color determination device 1 will be described with reference to specific examples. The flowchart of the two-dimensional colorimeter 5 is shown in FIG. 8, the flowchart of the IOT server 6 is shown in FIG. 9, and the flowchart of the illumination color conversion unit 7 is shown in FIG.

When the power of the two-dimensional colorimeter 5 on the local side is turned on, initialization is performed as shown in FIG. 8 (initialization S1). Next, the vehicle N1 is imaged by the spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)) (imaging process S2). The imaging process S2 is a step of imaging the vehicle N1 with a two-dimensional colorimeter 5 having three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) (see FIG. 6). The spectral sensitivity (S1 (λ), S2 (λ), S3 (λ)) is given according to Equation 1. The input process S3 is continuously performed at the same time as the image is taken by the photographing lens 51, the optical filters 52a, 52b, 52c and the image pickup element 53.

After that, the captured 3-band visual sensitivity images S1i, S2i, S3i (T = 6500K) are input by the image pickup element 53 (input processing S3), and the arithmetic processing unit 54 inputs the captured 3-band visual sensitivity images S1i, S2i, S3i ( T = 6500K) is transmitted to the IOT server 6 (transmission process S4), these are recorded on a recording medium such as a two-dimensional colorimeter 5 (recording process S5), and it is determined whether or not to end (S6), and the process is performed. To finish.

Next, when the power is turned on to the IOT server 6 on the remote side, initialization is performed as shown in FIG. 9 (S110). A plurality of 3-band visual sensitivity images S1i, S2i, S3i (T = 6500K) relating to a plurality of vehicles N1 or a standard plate N2 transmitted from the 2D colorimeter 5 are received, and the 3-band visual sensitivity images S1i, are stored in the storage unit 63. Record as S2i and S3i (S120). When the image is a moving image, a series of processes are continuously performed.

When IC tag T information (including vehicle type, color, texture (metallic information), etc.) is received from the control PC 8, a registered 3-band visual sensitivity image corresponding to the color / texture information included in the IC tag T information is received. S1i, S2i, and S3i are read out and transmitted to the illumination color conversion unit 7 (S130).

It is determined whether or not the registration request has been received (S140). If YES, the process proceeds to S150, and if NO, the process proceeds to the display process (S160).

Upon receiving the image registration request from the two-dimensional colorimeter 5, the three-band visual sensitivity images S1i, S2i, and S3i captured and recorded by the two-dimensional colorimeter 5 are newly registered or updated (S150).

The process related to registration is displayed on the display unit 64 of the IOT server 6 (S160).

It is determined whether or not to end the processing, and if the processing is not completed, the process returns to S120, and if the processing is completed, a return is obtained (S170).

The processing of the illumination color conversion unit 7 will be described with reference to FIG. The illumination color conversion unit 7 has a configuration independent of the IOT server 6, for example, a personal computer, but may be provided in the IOT server 6 and configured as a part thereof. The color temperature correction process using ambient light will be described.

The received 3-band visual sensitivity images S1i, S2i, S3i (T = 6500K) are temperature-corrected for each pixel in the arithmetic processing unit 54 of the two-dimensional colorimeter 5, converted into a tristimulus value XYZ, and XYZ-. The RGB image is transmitted to the wearable terminal 3 by RGB conversion. The XYZ conversion is performed by the formula 2 on the local side, but on the remote side where the environment is different from the local side, the conversion considering the correction is performed according to the formula 3 due to the temperature correction. As shown in Equation 3, the tristimulation value XYZ of the XYZ color faithful image is corrected to become the tristimulation value X', Y', Z', and the image from the local side is imaged on the remote side based on this correction value. Is displayed.

Here, the matrix with the D65 standard white light source (T = 6500K) so that the heads of the curves of the spectral sensitivities S1, S2, S3 appearing in the conversion formula of XYZ-S1, S2, S3 of the following formula 2 are aligned. Perform coefficient correction. For details, see IEICE TRANS.INF. & SYST., VOL.E93-D, No.3 MAR 2010 copy rght 2010 The Institute of Electronics, Information and CommunicationEngineers, Development of an XYZ Digital Camera with Enbedded Color Calibration System for Accurate. Please refer to Color Acquisition, Maciej KRETKOWSKI, Ryszard JABLONSKI, SHIMODAIRA P651-653 for white balance and color calibration.

The illumination color conversion correction unit 7 performs ambient light correction on the local side and the remote side. For example, in the environment on the local side, T = 6500K, and in the environment on the remote side, for example, the environment color temperature is assumed to be T = 5000K for explanation, but the color temperature is particularly relevant because it is actually measured by the micro color temperature measurement sensor 4. There is no such thing, but the purpose is to improve the accuracy of color / texture data by correcting for temperature differences. The configuration of the illumination color conversion correction unit 7 on the remote side and the color temperature measurement sensor 4 which can be connected to the remote side and is equipped with a small spectroscope such as a micro spectroscope is used. Hereinafter, description will be made with reference to FIGS. 10 to 12.

When the color temperature of an existing lighting measured by the color temperature measurement sensor 4 is used, their spectra do not always match (Color Engineering Noboru Ota (Tokyo Electric University Publishing Bureau) draws a color temperature line to color temperature. Even if it is not derived from the spectral sensitivity of S1, S2, S3 from the black body radiation from the color temperature, the three numerical values of the illumination light on the local side and the remote side are the spectra of S1, S2, S3. S1gain and S3gain can be calculated using three values obtained by multiplying the sensitivity by the spectra of the illumination measured on the local side and the remote side. Each of them can be selected arbitrarily on the remote side. It is characterized by obtaining the ratio of the spectral characteristic output of S123 of the local side illumination and the spectral characteristic output of S123 of the remote side illumination.

As shown in FIGS. 10 to 12, the illumination color conversion correction unit 7 has T = 6 for the three-band visual sensitivity images S1i, S2i, S3i (T = 6500K) of the vehicle N1 or the standard plate N2 recorded by the IOT server 6. The temperature is corrected from 6500K to T = 5000K, and the calculated image data is transmitted to the wearable terminal 3.

The processing performed by the illumination color conversion correction unit 7 will be described. In S200, the remote side receives the three-band visual sensitivity images S1i, S2i, S3i (T = 6500K) obtained by imaging the car 206 on the local side with the two-dimensional colorimeter 202 from the local side. On the local side, as shown in FIG. 11, when the output value of the illumination spectrum of 6500K and the spectral sensitivities S1, S2, S3 (see Equation 1 and FIG. 5) are multiplied, the multiplication values S1, S2, S3 as shown by diagonal lines. Is sought. Integrate these to calculate the integrated values IS1, IS2, IS3 within the curve range shown by the diagonal line, and adjust these integrated values, for example, the gain of the two-dimensional colorimeter 5 so that IS1 = IS2 = IS3. Is adjusted, and this is used as the standardization sensitivity of the two-dimensional color meter 5.

In S210, the color temperature measurement sensor 4 on the remote side calculates S1gain, S2gain (this is relatively set to 1), and S3gain at T = 5000K. As shown in FIG. 12, the output value of the illumination spectrum of T = 5000K on the remote side (5000K is an example and may be another temperature) is multiplied by the normalized sensitivities S1, S2, S3 (Equation 1, see FIG. 5). Then, these are integrated and the integrated values I03, I02, and I01 within the curve range shown by the diagonal line are calculated to obtain S1gain = I01 / I02 and S3gain = I03 / I02. This calculation simplifies the calculation of the gain value obtained by the next remote calculation. Even if you do not adjust the gain here, set S1gain = (IO1 / IO2) / (IS1 / IS2), S3gain = (IO3 / IO2) / (IS3 / IS2), which are the conversion formulas at that time, and then set the following. If the gain calculation on the remote side transmits the integrated value on the local side to the remote side, the calculation on the remote side becomes possible based on it. As described above, unlike the coefficient correction of the matrix described in Equation 2, it is not necessary to perform more operations for advance preparation for aligning the heads of S1, S2, and S3, so such gain adjustment is performed. By doing so, the S1, S2, S3-XYZ conversion formula of the above-mentioned formula 2 is associated with the lighting on the local side.

In S220, the 3-band visual sensitivity images S1i, S2i, and S3i (T = 6500K) acquired in S200 are corrected by S1gain and S3gain calculated by S210, and the 3-band visual sensitivity images S1i5000k and S2i5000k (same as S2i) are corrected. ), S3i5000k (T = 5000K) is derived. This is the same operation as in Equation 3.

In S230, the light incident portion (slit) of the color temperature measurement sensor 4 is pressed against the image displayed on the screen of the display unit 2, the image is captured, and the R data (for example, R = 255: 8 bits) value. Create a conversion table (may be a conversion matrix) for converting XYZ values to RGB values by linking G data with Xg, Yg, Zg, and B data with Xb, Yb, Zb. .. In this process, the ambient light on the remote side is taken into the illumination color conversion correction unit 7 via the color temperature measurement sensor 4, and the spectrum of the ambient light on the remote side is accurately reflected in the XYZ-RGB conversion by the color temperature measurement sensor 4. The purpose is to display each RGB data on the color temperature measurement sensor 4 and acquire a spectrum for each. As a result, X _R , Y _R , Z _R for R can be obtained, so if G and B are obtained in the same way, X _G , Y _G , Z _G for G and X _B , Y _B , Z for B. _B asks. In this way, the conversion table from XYZ to RGB can be obtained. This step 230 may be processed in advance and programmed.

In S240, the accurate 5000k of each image is obtained from S1i5000k, S2i5000k, S3i5000k (i = 1 to m, m is the number of pixels) obtained by multiplying the spectral relative gains S1gain, S2gain, and S3gain (T = 5000K) obtained in S220. The XYZ value (T = 5000K) is obtained. It is calculated by the above-mentioned formula 2.

In S250, the corresponding RGB value (T = 5000K) is obtained by using the XYZ-RGB conversion table for the XYZ value (T = 5000K) obtained in S240. By the above color temperature correction operation, the image can be accurately displayed on the RGB monitor in response to changes in the environment.

In S260, the RGB value (T = 5000K) obtained in S250 is transmitted to the wearable terminal 3, and the image corresponding to the RGB value is displayed on the display unit 2 of the wearable terminal 3.

Next, the color determination method according to the first embodiment will be described with reference to FIGS. 1 to 12. Since the detailed description of the configuration and operation has been described above, it will be referred to and described according to the methodical elements.

First, a plurality of 3-band visual sensitivity images S1i, S2i, S3i obtained by imaging a plurality of vehicles N1 or a standard plate N2 under a standard light source are transmitted to the IOT server 6 by the two-dimensional colorimeter 5 and stored (memory step). ).

The storage step stores the tag information of the product, and stores the three-band visual sensitivity images S1i, S2i, and S3i of the standard plate corresponding to the tag information (tag information storage step).

The color temperature is measured by the color temperature measuring sensor 4 under the remote illumination light (color temperature measuring step).

The three-band visual sensitivity images S1i, S2i, and S3i are converted into XYZ values with color temperature correction based on the measured values of the color temperature measuring sensor 4, further converted into RGB values, and transmitted to the wearable terminal 3. (Illumination color conversion correction step).

An image corresponding to the RGB value is displayed on the display unit 2 of the wearable terminal 3 (display step).

As shown in FIG. 13, in the body assembly process, face-to-face is performed at the time of assembly, so the color and texture are carefully examined. The probability can be significantly reduced.

According to the embodiment described above, the standard color and texture are displayed on the display unit 2 of the wearable terminal 3 by simply attaching the wearable terminal 3 to a person, and the other work flow is shown by comparing with the actual product. It is possible to dramatically improve the skill level of the color / texture determination work by the vehicle N1 and the standard plate N2 without hindering. Compared with the conventional color / texture inspection, the present invention has a high quality advantage because the illumination color conversion correction unit 7 corrects the color / texture environment, and the color / texture on the local side and the remote side. It can be shared. By using the wearable terminal 3 to make it a wearable terminal 3 that incorporates measurement technology and information processing with a two-dimensional colorimeter, it can be used as an alternative tool to the standard board. There is an advantage that even low-ranking workers can immediately perform advanced inspections related to color / texture judgment.

It is possible to reduce costs by improving the efficiency of the color / texture matching process and reducing the waste of color / texture correction. In addition, the introduction cost is low, and it is highly accurate and stable.

An example of application to the process of assembling the bumper B to a vehicle in a finished vehicle factory has been given, but it goes without saying that it can also be applied to inspection in a parts factory as shown in FIG.

It enables quick unification of the color and texture of the combined body. With the IOT technology, the color and texture of the car being carried is described in the IC tag T etc. attached to the body, so the color and texture of the corresponding standard board is automatically displayed in the wearable terminal 3. Will be done. In this way, in addition to checking the color and texture in each process, it is possible to check with other parts and standard plates at the time of assembly, so the probability of defective discharge is further reduced.

The color determination device 101 of the second embodiment will be described. The color determination device 101 has a part number in the 100s for the configuration common to the color determination device 1 of the first embodiment, and the above description and illustration are used to mainly explain the differences. The wearable terminal 103 of the color determination device 101 is provided with a small RGB camera 135. The RGB small camera 135 measures the color distribution of the painted surface of the bumper B, and displays the color and the feeling of spread of the distribution separately. The IOT server 106 transmits information on both the corrected XYZ value and the RGB value of the standard plate N2 to the wearable terminal 103. This is a method of quantifying the metallic feeling and comparing it with the standard plate N2. The inspection is performed using the wearable terminal 103, but if there is no difference with the human eye, there is no problem. However, if the human eye feels a difference in color and texture, the result is displayed by comparing this difference in color and metallic feeling with the data of the standard plate measured locally. In this case, the display 102 is displayed by pressing down the switch.

As shown in FIG. 15, the wearable terminal 103 includes a display 102, a small camera 135, a touch pad 136, a speaker 137, and a controller 138 in a glasses main body 134 having a lens, as shown in FIG. In one example, the display 102 may include an image generation element disposed within the lens. As another example, the display 102 may include an optical modulator at the edge of the lens.

The processing performed by the controller 138 will be described below with reference to FIGS. 16 to 27.

When the power of the wearable terminal 103 is turned on, the wearable terminal 103 is initialized as shown in FIG. 16 (initialization S1). Next, the bumper B is imaged by the small camera 135 (imaging process S2), then the captured image data is input from the small camera 135 to the controller 138 (input process S3), and the controller 138 stimulates the RGB values. It is converted into values X, Y, Z and converted into xy (conversion process S4). When the image is a moving image, a series of processes from the image pickup process S2 to the data transmission S5 are continuously performed. Please refer to Japanese Patent Application Laid-Open No. 2016-6416 for RGB-XYZ conversion.

In the image pickup process S2, the bumper B is imaged by the small camera 135 at different angles in the specific areas where the image pickup positions are different. There are multiple imaging points, and the number can be selected as appropriate. Here, measurement is performed from three directions of front (0 degree), left 45 degrees, and right 45 degrees. As for the measurement location, the 0 degree optical axis of the small camera 135 is perpendicular to the body surface of the bumper B. In addition, the lighting is the lighting from diagonally above like the sunlight.

The conversion formulas from the three stimulus values X, Y, Z to the Y'xy color system are given in Equations 4 and 5. Here, a luminance meter (not shown) is used together with the small camera 135, and Y is calibrated by the value (nt) of the luminance meter to be Y'. Since the color space conversion formula is a conventional one, other detailed formulas are omitted.

The XYZ color system is currently the basis of each color system as the CIE standard color system. It was developed based on the principle of additive color mixing of the three primary colors of light (R = red, G = green, B = bluish purple), and the color is represented by the three values of Yxy using a chromaticity diagram. Y corresponds to the lightness by the reflectance, and xy becomes the chromaticity.

The sub-flow chart of FIG. 17 will be described. This arithmetic processing is a step of calculating and visualizing the Lab average value, ΔL, Δa, Δb, ΔE and xy texture spread index of the standard plate N2 and the captured image, and is necessary for displaying on the display 102. In that case, the color information is converted into RGB or the like. The first image of the standard plate N2 is imaged, the second image of the inspection object to be compared next is imaged, and the texture spread index is sequentially calculated as follows. The texture similarity is determined by the texture spread index that separates the textures.

The inspection area K (see FIG. 21B) corresponding to the region T (see FIG. 19A) to be inspected for the captured images A and B is set (step S441). The size and location can be set freely.

The chromaticity xy is calculated to obtain the chromaticity Yxy (S442).

An xy chromaticity histogram distribution of the region K cut out from the image A of the standard plate N2 is created (S443). This chromaticity histogram distribution is an integrated number obtained by counting the pixels belonging to the overlapping region D of the two histogram distributions shown in FIG. 19 (c).

The xy chromaticity histogram distribution is a three-dimensional histogram showing the integrated number of pixels belonging to each unit cell, and shows an overlapping region D in FIG. 19 (d).

As shown in FIG. 19 (c), the color distribution of the comparison target at the position of the xy coordinates is drawn in a plane, the inspection area K is divided by the grid G, and the pixels having the xy value of the division are integrated. Create a histogram distribution with the z-axis. Let the xy coordinates be a grid (three-dimensional grid) of a specific width, for example, a planar grid obtained by cutting xy by 1/1000 (1000 lines). The histogram is scanned from one end to the other, and the number of pixels belonging to each region divided by the grid G is scanned on the same xy plane and integrated in the z direction. Further, if only the vertical plane or the horizontal plane in a specific range is calculated in the inspection area K with the xy coordinates, the calculation time can be shortened. If the grid is made finer, the accuracy will be improved, but the calculation time will be longer, so use appropriate grids.

Similar to S443, the xy chromaticity histogram distribution of the image B captured by the small camera 135 is created (S444). The xy chromaticity histogram distribution is the integrated number of pixels, and shows an overlapping region in FIG. 19 (d).

For the a-axis, b-axis, and L-axis of Lab, the sum of all the pixels in the inspection area is taken independently, and the sum of each L value, a value, and b value is divided by the number of pixels to obtain the Lab color. The average L value, the average a value, and the average b value of the spatial histogram distribution are calculated (S445).

The Lab value of the Lab space converted by the following formula 4 is calculated. The Lab color space is a kind of complementary color space, has a dimension L meaning lightness and complementary color dimensions A and B, and is based on a non-linear compression of the coordinates of the CIEXYZ color space. By converting the XYZ value before normalization into Lab by the mathematical formula 6, a distribution in the Lab color space that takes into account the brightness direction can be obtained as opposed to the distribution in the XYZ color space.

In Equation 6, the values of X, Y, and Z in parentheses of the function f are divided by the coordinates X _n , Y _n , and Z _n of the white point, respectively, in order to align the maximum values with 1.

The difference between the average values of the images A and B is taken and used as a judgment material for the difference in color and texture.

As shown in FIG. 18, the center coordinates C ₁ and C ₂ of the xy color space histogram distribution are specified (S446). Here, the center coordinate is the center of gravity (center of gravity position).

As shown in FIG. 18, the deviation ΔF of the center coordinates so that the center coordinates of one of the two xy histogram distributions H ₁ (x, y) and H ₂ (x, y) match the other center coordinates. Only, the entire xy chromaticity distribution is shifted (mapped) (S447). This is because if one of the distributions is not shifted to the other distribution, the difference in color components will also be calculated. You can do it on the graph or just by calculating. The shift amount can be set as appropriate. For example, instead of shifting from one center to the other center, a shift from one center to the other center within a predetermined range has the same effect. In short, it suffices to approach them with an appropriate shift amount that allows the texture to be grasped.

The texture spread index indicating the difference in spatial spread is calculated (S448). This makes it possible to simply extract only the difference in metallic feeling, determine the similarity of chromaticity and the degree of metallic feeling separately, and quantify this. The degree of spread is calculated in the two-dimensional space of the xy chromaticity distribution, and the difference in the degree of spread can be grasped as the difference in the glitter feeling of the glitter material excluding the color, so the color and texture can be surely understood. Can be detected separately.

The texture spread index is calculated by the following formula. The xy chromaticity histogram distribution is the integrated number of pixels, and the overlapping region D is shown in FIG. 19 (d), and the minimum distribution is shown in FIG. 19 (e).
Texture spread index = integrated number of pixels belonging to overlapping area D / total number of pixels in inspection area K x 100 (%)

The spreadness histogram of the standard plate N2 and bumper B in the two-dimensional space is calculated, and the minimum value of the same positions in the array is taken as the overlap frequency. It is calculated by dividing by.

19 (d) and 19 (e) are one cross-sectional view of FIG. 19 (c) cut out from the SS cross section, and there is an overlap when viewed on the same line in xy coordinates. Instead of drawing three-dimensionally, it is drawn on a plane for convenience. Moreover, since it is a histogram, it has a minute stepped distribution. The integrated number H1 and the integrated _number H2 in _FIG . 19D correspond to the image A and the image B, respectively. Comparing the two histogram distributions, there is an overlapping region D.

As shown in FIG. 19 (e), H ₁ (x ₁ , y ₁ ) is the integrated number of the xy chromaticity histogram distribution of the standard plate N2, and H ₁ (x ₂ , y ₂ ) is the xy chromaticity histogram distribution of the bumper B. In the overlapping left region, H ₁ > H ₂ , H ₁ = H ₂ in the center, and H ₁ <H ₂ on the right side. Taking the smaller integrated number (pixel frequency) of H ₁ and H ₂ , it becomes H ₁ on the left side and H ₂ on the right side, and the minimum distribution which is a stepped histogram curve can be specified. Using this, the ratio of the overlapping area D to the entire area can be calculated.

This minimum distribution identifies the smaller integrated value. By adding the smaller integrated number of H ₁ and H ₂ , the integrated number of the overlapping region D can be calculated, and the ratio to the total number of pixels can be specified. The total number of pixels in the inspection area K is fixed, and the total number of pixels is the same in both the bumper B and the standard plate N2. The calculation of this ratio may be three-dimensionally integrated for all the grids G. For example, as shown in FIG. 19 (c), the inspection region K is cut along the SS axis, and y is a predetermined value. The distribution of the integrated number of pixels when x changes from end to end is integrated two-dimensionally. FIG. 19 (f) is a two-dimensional map of the integration result on the xy coordinates. If there is no distribution in the inspection area K and the number of pixels is zero, it is excluded from the calculation.

Finally, display / save processing such as displaying the visualized texture spread index and the like, transmission processing (S449) are performed, and the processing is returned.

For example, the pixels belonging to the inspection area K are 100 vertical pixels x 100 horizontal pixels = 10,000 pixels. Since the image is cut out in the same inspection area K, the total number of pixels of the image A and the image B is 10,000 pixels. From the xy chromaticity histogram, the number of pixels in the overlapping region is integrated, and when the integrated number is 5,000, the texture spread index is 50%. The degree of difference in texture increases as the texture spread index falls below 100%. If the distribution of xy values is completely the same, it will be 100%. As a result, when it is determined that the numerical value is equal to or higher than a certain value, it can be determined that the texture is a conforming product.

For images A and B, the primary color information obtained is RGB because it has three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) by a function equivalent to the XYZ color matching function. It is more faithful to the sensitivity of the human eye and more accurate than the case of acquiring by. Since the overlap of the spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) is small, the S / N ratio is sufficient, and the curve in the spectral sensitivity curve changes naturally, the error in color measurement is It is kept to a minimum.

Since the texture of the image can be grasped by the histogram distribution separately from the color, the difference in color texture (metallic feeling, glitter feeling, mottled pattern, color pattern, rugged feeling, etc.) can be reflected to make a slight difference in hue. Can be determined.

As shown in FIGS. 20 (a) to 20 (c), an example in which three types of inspections are inspected, from those having a small degree of metallicity to those having a large degree of metallicity, will be described. The one with a small degree of metallicity is referred to as the standard plate N2, the one with a medium degree of metallicity is referred to as bumper B1, and the one with a high degree of metallicity is referred to as bumper 2. First, when the distribution on the xyz chromaticity diagram after performing the above processing is created, the highlight portions are integrated data as shown in the xy chromaticity diagram of FIG. 20B. The total number is shown in light and dark, and the brighter the color, the larger the total number. FIG. 20 (c) schematically shows the integrated number in three dimensions of the standard plate N2 and the inspection object. The xy axis is the chromaticity, and the z axis is the integrated number. Basically, the stronger the metallic feeling, the lower and wider the chevron shape, and the weaker the metallic feeling, the sharper the chevron shape. When the bright material (aluminum flakes), which is the source of the metallic feeling, gives a glittering feeling due to minute protrusions and the like when exposed to illumination light, this glittering feeling is physically a diffraction phenomenon of light. By comparing the two histogram distributions of the standard plate N2 and the inspection target 2 or 3, the texture spread index indicating the degree of overlap is calculated.

As shown in Table 1, since the comparative example uses a Lab in which ΔE calculates the color that is the source of the texture as an average value, the Lab value and the ΔE value are minute differences compared to the appearance, and it is difficult to inspect. there were. Since the integrated number within the range of the inspection area K is used as it is for the texture spread index of the present embodiment, the inspection objects 2 and 3 are 58% and 27%, respectively, with respect to the standard plate N2, which are clearly and numerically. , The metallic degree can be easily identified.

The next color determination device 101 of the third embodiment will be described with reference to FIGS. 21 and 22. For the corresponding similar elements, the explanation is referred to as the 200 series, and the differences are mainly explained.

A controller 238 that is connected to a small camera 235 and a small camera 235 that captures the bumper B to receive a signal and calculates a texture spread index, and a display 202 that is connected to the controller 103 to display exponential notation and matching degree. I have.

As shown in FIG. 21, the controller 103 has a calculation unit 203A for recording XYZ1 which is a corrected XYZ value for receiving XYZ1 corresponding to the standard plate N2 from the IOT server 206, and XYZ2 acquired by imaging the bumper B. The calculation unit 203B for calculating the calculation unit 203B, the calculation unit 203C for calculating the color / texture matching degree index of the car by connecting the calculation unit 203A and the calculation unit 203B, and the OK signal or NG signal, the index display, etc. from the calculation unit 203C. It is transmitted to the display 202 for display or transmitted to the outside.

FIG. 22 is a flowchart for calculating the texture spread index by comparing the chromaticity histogram distributions from the two images A and B. As shown in FIG. 11, when the program is started, the inspection area K is cut out from the image A, specified, and set (S501). Next, an inspection area similar to that of image A is cut out from image B, specified, and set (S502). The chromaticity value XYZ is calculated from the images A and B (S503). In the inspection area K, the XYZ chromaticity histogram distributions of the bumper B and the standard plate N2 are calculated and created (S504). The average value of the XYZ values is calculated (S505). The center coordinates of the XYZ color space distribution are specified (S506). The XYZ color space distribution to the center coordinates is shifted (S507). After the shift process, the center coordinates of the XYZ color space distribution are specified (S508). This is to confirm the suitability of the center coordinates after the shift process. The center coordinates can be readjusted here. The minimum distribution of the XYZ chromaticity histogram distribution is specified, and the integrated number of the XYZ chromaticity histogram distribution in the overlapping region D is calculated (S509). Texture spread index = (integrated number of pixels belonging to the overlapping region D / total number of pixels in the inspection region K) × 100 (%). For the integrated number in the overlapping region D, the integrated number of T ₁ and T ₂ whichever is smaller is added. Calculate the texture spread index (S510) and return.

In the case of the calculation of the XYZ distribution corresponding to the inspection area K, the calculation of the exponent is performed by the distribution in the three-dimensional space of the X-axis, the Y-axis, and the Z-axis. The XYZ values in XYZ space coordinates of the bumper B and the standard plate N1 are T ₁ (L, a, b) and T ₂ (L, a, b), respectively, as shown in FIGS. 24 (a) and 24 (b). And. In the XYZ color space, the histogram distribution has a shape like a globe, and there are cases where the two histogram distributions are sterically overlapped and separated. This is shifted and the center coordinates are brought closer. The inspection area K in the three-dimensional space is divided by a grid, and the chromaticity histogram distribution and the minimum distribution of T ₁ (X, Y, Z) and T ₂ (X, Y, Z) in three dimensions are obtained, and the same exponent is obtained. Perform the operation of. The integrated number of the grid may be projected on a plane, and the integrated number of the overlapping region on the grid may be calculated by the same integration in the plane. In the case of XYZ chromaticity, there is no brightness information, so in the XYZ space, the histogram distribution does not change even if the brightness of the image changes.

When the Lab color space histogram is used for the texture determination instead of the XYZ color space histogram, the flowchart of FIG. 23 is used. The description of FIG. 23 is based on the above description of FIG. 22. In S505, the average Lab value of the region is calculated and the average Lab value of the inspection region of the image B is calculated. In the case of Lab chromaticity, since there is brightness information, in the Lab space, the histogram distribution changes when the brightness of the image changes.

Next, the color determination device 301 of the fourth embodiment will be described with reference to FIG. 25. For the corresponding similar elements, the explanation is referred to as the 300 series, and the differences are mainly explained.

As shown in FIG. 25, the color determination target is a part of the color sample 5, and the small camera 335 captures the target area of the bumper B. The controller 338 is connected to a calculation unit 303A that calculates a Lab from the corrected XYZ1 of the standard plate N2, a calculation unit 303B that calculates a Lab from the XYZ2 of the bumper B to be determined, and a calculation unit 303A and a calculation unit 303B. A calculation unit 303C that calculates the Lab average value, a texture spread index calculation unit 303D that calculates the texture spread index from the standard Lab and the target Lab, and a device that transmits and displays the calculation values from the calculation units 303C and 303D to the display 202. Is. Depending on the index value, it can be determined by looking at the screen whether or not the paint color and texture are appropriate. Since the main processing is almost the same as the flowcharts of the second and third embodiments, the description is incorporated.

In the case of the calculation of the chromaticity histogram distribution in the Lab space corresponding to the inspection area K, the XYZ value is converted to the Lab. The calculation of the exponent is performed by the distribution in the three-dimensional space of the L-axis, the a-axis, and the b-axis. The Lab color space histogram distribution is a three-dimensional elliptical shape. The Lab values in Lab space coordinates of the bumper B and the standard plate N2 are U ₁ (L, a, b) and U ₂ (L, a, b), respectively, as shown in FIGS. 26 (a) and 26 (b). And. In the color space of Lab, the histogram distribution has a shape like a globe, and there are cases where the two histogram distributions are sterically overlapped and separated. The inspection area K in the three-dimensional space is divided by a grid, and the chromaticity histogram distribution and the minimum distribution of U ₁ (L, a, b) and U ₂ (L, a, b) in three dimensions are obtained, and the same exponent is obtained. Perform the operation of. The integrated number of the grid is projected on a plane, and the integrated number of the overlapping region on the grid is calculated by the same integration in the plane. In the case of Lab chromaticity, since there is brightness information, in Lab space, when the brightness of the image changes, the L value changes, and the distributions U1 and _U2 of the degree of agreement _are positioned in Lab space. Therefore, it is possible to make a judgment in consideration of lightness and darkness. This is because the position of the distribution shifts if the brightness of the image is different. For example, the Lab color space histogram distribution shifts downward when it gets dark, and shifts upward when it gets bright.

Further, the process by the color determination device according to the fifth embodiment of the present invention is an example of determining the gloss and shine of the skin in the Lab color space histogram distribution based on the difference in distribution, as shown in FIG. 27.

In the case of xy chromaticity histogram distribution, color is judged by ΔL, Δa, Δb value and ΔE (√ (ΔL ** 2 + Δa ** 2 + Δb ** 2) (** 2 means square). For texture, With a two-dimensional distribution of normalized chromaticity values x and y that do not consider the L value, the texture spread index is quantified only by the color spread distribution. In the case of the XYZ color space histogram distribution, each X value of XYZ, Color is judged by the average value of Y value and Z value (color difference judgment by Lab value is also OK). Texture is judged by the texture spread index only by the distribution of XYZ color space histogram. In the case of Lab color space histogram distribution, ΔL , Δa, Δb values and ΔE (√ (ΔL ** 2 + Δa ** 2 + Δb ** 2)). For the texture, the texture is determined by the texture spread index based only on the distribution of the Lab color space histogram.

Other application examples will be described. The two images of the standard plate N2 and the acquired A and B images of the bumper B are superimposed, and the respective chromaticity histogram distribution or color space histogram is displayed on the display 102, or the respective chromaticity histogram distribution or color space is displayed. A chromaticity diagram in which the histograms are superimposed on one chromaticity diagram can be displayed, and the difference in color can be determined by the average Lab value, while the texture spread index calculation showing a metallic feeling can be separately displayed as a percentage. As a result, the deviation of the texture, especially the metallic feeling, can be surely confirmed numerically from the distribution of the bumper B and the standard plate N2. The inspection result is displayed numerically for each area K. The grid width of the grid can be adjusted. The index threshold can be set arbitrarily. The measurement results and captured images can be saved. It is possible to standardize colors and textures and perform stable color and texture management by reducing problems of individual differences that cannot be avoided by visual inspection and troubles of judgment criteria with customers.

Since non-contact and wide-range photography is possible, flip-flops can be quantified by imaging the standard plate N1 from multiple angles with flat lighting, such as painting with aluminum flakes and pearl pigments. It is possible to evaluate the color and texture that the eyes feel. Irregularly patterned parts such as wood grain panels can also be matched in color and texture. Since the captured images A and B can be displayed on the displays 102, 202, and 302 (overlay function), the alignment can be easily performed. The inspection object can be compared with the standard plate N2 even if the size and material are different. Fabrics with irregular patterns and textures such as leather can also be matched in color and texture. It is possible to inspect resin parts and color unevenness / color deviation. For example, it is possible to measure an object having irregularities. You can also color match building materials with irregular patterns and textures such as flooring, irregular patterns such as wallpaper, and textures such as wood grain, marble style, and geometric patterns. You can inspect the texture of teeth in the field of dentistry.

Although the present embodiment has been described above, there are the following effects in addition to the same as the first embodiment. Compare the bumper B and the standard plate N2 with the human eye on the wearable terminals 103, 203, 303, and if you feel a difference in color / metallic feeling, press the switch as appropriate to see the difference in color / metallic feeling on the local side. The displayed result can be displayed by comparing with the measured data of the standard plate N2, the degree of agreement can be accurately recognized, and the pass / fail can be determined. This makes it possible to speed up and improve the determination work.

It should be noted that the embodiments of the present invention are not limited to the above-described embodiments, and modifications and the like can be added within a range that does not deviate from the technical idea of the present invention. It goes without saying that objects and the like are also included in the technical scope of the present invention and can take various forms as long as they belong to the technical scope. For example, the wearable terminal is not limited to these, and the technical idea of the present invention is also implemented by other types of wearable terminals. For example, regarding the method of acquiring image information according to the three-band visual sensitivity images S1i, S2i, and S3i, the method given in the present embodiment is only a specific example. Although the arithmetic processing of the second embodiment is performed by the controller 138, the operation may be performed by the IOT server 6 or the control PC 8 to transmit the signal to the controller 138.

The color determination device of the present invention is an accurate electronic color / texture inspection using a standard plate at a manufacturing site of a standard product that serves as a standard for color / texture or an electric appliance, a vehicle, a housing building material, etc. that uses a color standard plate. , Other industrial applicability is great. Inspection of printed matter color deviation / color unevenness, inspection of cosmetics containing lame / pearl pigment, automobile painting inspection (color deviation / color unevenness, color deviation between different materials are quantified, flip floc by graphing light brightness (Quantification of phenomenon), wood grain inspection of automobiles, leather seat inspection of automobiles (inspecting color unevenness and color deviation of complicated texture materials), color matching of floor materials, color matching of tiles (accurately in color and texture) (Photographing), color matching inspection of cosmetics, sportswear, shoes, golf balls, etc.

1 ... Color determination device 2, 2'... RGB display unit 3, 3'... Wearable terminal 4 ... Color temperature measurement sensor 5 ... Two-dimensional colorimeter 6 ... IOT server 7.・ ・ Illumination color conversion correction unit 8 ・ ・ ・ Control PC
9 ... Micro optical system reading spectroscope 10 ... System control unit 11 ... Color correction unit 12, 63 ... Storage unit 13 ... Signal pattern generator 15 ... High color performance monitor 31 ...・ Arm band 32 ・ ・ ・ Head band 33 ・ ・ ・ Control box 51 ・ ・ ・ Shooting lens 52a ′ 52c ′ ・ ・ ・ Dichroic mirror 52a, 52b, 52c ・ ・ ・ Optical filter 53 ・ ・ ・ Imaging element 53a, 53b, 53c ・ ・ ・ Imaging element 54 ・ ・ ・ Calculation processing unit 55 ・ ・ ・ Image display unit 56 ・ ・ ・ Reflector 57 ・ ・ ・ Filter turret 61 ・ ・ ・ Input / output interface 62 ・ ・ ・ Calculation unit 64 ・ ・ ・ Display Part 65 ... Bus line N1 ... Car N2 ... Standard plate S1, S2, S3 ... Spectral sensitivity

Claims (2)

A wearable terminal on the remote side that has a display unit that displays an image, receives a signal, and displays the image on the display unit.
A color temperature measurement sensor on the remote side that measures the color temperature under remote illumination light,
Multiple color standards under a standard light source by a two-dimensional colorimeter with three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) linearly transformed to the CIE XYZ color matching function. It is equipped with a local IOT server that stores a plurality of 3-band visual sensitivity images S1i, S2i, and S3i that have been imaged.
The wearable terminal is provided with an arithmetic unit and a small camera, and the arithmetic unit can be used as an arithmetic unit.
The three-band visual sensitivity images S1i, S2i, and S3i stored in the IOT server are adjusted for the gain of the two-dimensional colorimeter based on the integrated value of the measured values of the color temperature measuring sensor, and the color temperature is corrected. It has an illumination color conversion correction unit that converts to the XYZ correction value made.
The IOT server stores the tag information of the product to be color-determined, and the wearable terminal receives the XYZ correction value of the color standard corresponding to the tag information from the illumination color conversion correction unit . A corrected image corresponding to the XYZ correction value is displayed on the display unit, and the corrected image is displayed.
The small camera of the wearable terminal captures an image of the product to be color-determined, the arithmetic unit measures the color distribution of the captured product image, compares the corrected image with the product image, and obtains the color. The texture spread index is calculated separately, and the comparison result is displayed on the display unit.
A color determination device capable of displaying a chromaticity diagram in which a chromaticity histogram distribution or a color space histogram of each of the product image and the correction image is superimposed on the chromaticity diagram on the display unit. Multiple color standards under a standard light source by a two-dimensional colorimeter with three spectral sensitivities (S1 (λ), S2 (λ), S3 (λ)) linearly transformed to the CIE XYZ color matching function. A storage step for storing a plurality of 3-band visual sensitivity images S1i, S2i, S3i captured in
A color temperature measurement step that measures the color temperature with a color temperature measurement sensor under remote illumination light,
The arithmetic unit of the wearable terminal adjusts the gain of the two-dimensional colorimeter based on the integrated value of the measured values of the color temperature measuring sensor for the three-band visual sensitivity images S1i, S2i, and S3i, and corrects the color temperature. Illumination color conversion correction step to convert to the XYZ correction value made
The IOT server stores the tag information of the product to be color-determined, the wearable terminal receives the XYZ correction value of the color standard corresponding to the tag information , and the wearable terminal displays the above. Display the corrected image corresponding to the XYZ correction value,
The small camera of the wearable terminal captures an image of the product to be color-determined, the arithmetic unit measures the color distribution of the captured product image, compares the corrected image with the product image, and obtains the color. The texture spread index is calculated separately, and the comparison result is displayed on the display unit.
A display step of displaying a chromaticity diagram in which the chromaticity histogram distribution or the color space histogram of each of the product image and the correction image is superimposed on the chromaticity diagram is displayed on the display unit.
A color determination method characterized by being equipped with.

Images