JP2002286549A - Method and system for inspecting color tone - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow detail color tone measurement, and allow color tone inspection for an inspection object having a pattern of plural colors. SOLUTION: The present invention provides a color tone inspection method for inspecting a color tone of a surface by an image processor 4 using a spectroscopic camera 1 for detecting a spectroscopic spectral value in each point in the surface of the inspection object having the pattern of plural colors, and a system therefor. The image processor 4 calculates a color tone data of each picture element corresponding to the each point of the surface by a color tone data image generating function 41 based on the spectroscopic spectral value in the each point in the surface, generates a color tone data image comprising the color tone data of the each picture element, partitions the color tone data image into plural divided areas by a partitioning function 42, conducts statistical calculation as to the color tone data of the each picture element in each of the plural divided areas by a statistical calculation function 43, compares an obtained statistically-calculated value with a prescribed inspection reference value by a comparison function 44, and inspects the tone color of the surface of the inspection object in response to a comparison result therein by a color tone inspection function 45.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color tone inspection method and system for inspecting the color tone of the surface of an inspection object having a plurality of color patterns (for example, color patterns and patterns).

[0002]

2. Description of the Related Art Conventionally, there have been two main methods for performing this kind of color tone inspection: a method using a color camera and a method using a spectrophotometer.

Japanese Patent Application Laid-Open No. 10-311756 discloses that a color unevenness of a colored plastic molded product is converted into a digital image showing the shading of the coloring, and the gradation value and its gradation value are assigned to each pixel constituting the image. The image is reproduced with the difference from the average of the gradation values in the vicinity area around the pixel as the gradation value of the pixel,
A method of quantifying color unevenness using statistical values of a gradation value distribution for all pixels of the reproduced image is described.

[0004]

However, the color tone inspection method using a color camera is not suitable for detailed color tone measurement because the color tone determination accuracy of the color camera is inferior. In addition, although the color tone inspection method performed using a spectrophotometer has excellent color tone discrimination accuracy of the spectrophotometer, it can only measure the average color tone of a predetermined area, so a pattern of a plurality of colors, for example, A color tone inspection of an inspection object having a complicated color pattern or pattern cannot be performed.

The present invention has been made in view of the above circumstances, and provides a color tone inspection method and a color tone inspection method capable of performing detailed color tone measurement and inspecting the color tone of an inspection object having a pattern of a plurality of colors. The purpose is to do.

[0006]

According to a first aspect of the present invention, there is provided an image processing apparatus, comprising: a spectroscopic sensor for detecting a spectral value at each point on a surface of an inspection object having a pattern of a plurality of colors; An image processing apparatus, which is a color tone inspection method for inspecting the color tone of the surface of the inspection object, calculating color tone data of each pixel corresponding to each point on the surface from a spectral spectrum value at each point on the surface. Generating a tone data image composed of tone data of each of the pixels; dividing the tone data image into a plurality of divided regions; and performing a statistical calculation on the tone data of each pixel in each of the plurality of divided regions. Performing a comparison, a step of comparing the statistical calculation value obtained in this step with a predetermined inspection reference value, and a table of the inspection object according to the comparison result. Characterized by a step of the color tone test.

In this method, a color tone data image is generated from the spectral spectrum values at each point on the surface of the inspection object, and the color tone data image is divided into a plurality of divided areas.
A statistical calculation is performed on the color tone data of each pixel in each of the plurality of divided regions, a comparison between the statistical calculation value and a predetermined inspection reference value is performed, and the color tone of the surface of the inspection object is determined in accordance with the comparison result. The inspection will be performed. As a result, a high-precision spectral spectrum value can be obtained from the spectroscopic sensor, so that high-precision color tone measurement can be performed. In addition, since the statistical calculation is performed in units of divided regions, there is no problem with visual inspection without being affected by fine patterns. Can be detected stably, and a color tone inspection of an inspection object having a pattern of a plurality of colors becomes possible.

According to a second aspect of the present invention, in the color tone inspection method according to the first aspect, the color tone data image is two-dimensional image data, and in the step of dividing into the plurality of divided areas, the generated color tone data The image is divided into a plurality of regions along at least one side and is divided into the plurality of divided regions. According to this method, it is possible to stably detect a slightly larger color unevenness, which is a problem visually, with relatively simple processing and without being affected by fine patterns.

According to a third aspect of the present invention, in the color tone inspection method according to the first aspect, the pattern of the pattern of the inspection object has a repetitive regularity, and the color tone data image is two-dimensional image data. In the step of dividing into the plurality of divided regions, the generated tone data image is divided so that a plurality of regions are arranged in a direction orthogonal to a direction in which the symbols having the repetition regularity in the tone data image are arranged. Then, the image is divided into the plurality of divided areas. According to this method, for example, it is easy to detect a color tone abnormality related to the regularity of the pattern of the lattice pattern.

According to a fourth aspect of the present invention, in the color tone inspection method according to the first aspect, the pattern of the pattern of the inspection object has a regularity of repetition, and the color tone data image is two-dimensional image data. In the step of dividing into the plurality of divided regions, the generated tone data image is divided so that a plurality of regions are arranged along a direction in which the symbols having the repetition regularity in the tone data image are arranged. It is characterized in that it is divided into a plurality of divided areas. According to this method, for example, it is easy to detect a color tone abnormality that is less relevant to the regularity of the pattern of the lattice pattern.

According to a fifth aspect of the present invention, in the color tone inspection method according to the first aspect, the inspection object is colored continuously from one end to another by a coloring device, and the color tone data image is Two-dimensional image data,
In the step of dividing into the plurality of divided regions, the generated tone data image is divided so that a plurality of regions are arranged along a direction orthogonal to a direction in which the coloring device continuously colors the image, It is characterized in that it is divided into a plurality of divided areas. According to this method, it is easy to inspect the continuous stability of the coloring apparatus, such as paint fading.

According to a sixth aspect of the present invention, in the color tone inspection method according to the first aspect, the inspection object is colored continuously from one end to another by a coloring device, and the color tone data image is Two-dimensional image data,
In the step of dividing into the plurality of divided regions, the generated tone data image is divided so that a plurality of regions are arranged along a direction in which the coloring device continuously colors, and the plurality of divided regions are divided. It is characterized by being classified into. According to this method, it is easy to inspect the color balance and the color unevenness in the coloring width direction of the coloring device.

According to a seventh aspect of the present invention, in the color tone inspection method according to the first aspect, in the step of dividing into the plurality of divided areas, the generated tone data image is divided by a division unit area serving as a minimum reference. Divided into multiple minimum areas,
A statistical distribution is obtained for the color tone data for each of the minimum areas, and each of the plurality of minimum areas is classified into a set where the value of the statistical distribution falls within at least one predetermined distribution range and a set which does not fall therein. Is divided into the divided regions. According to this method, it is possible to divide the object into a plurality of divided areas in accordance with the pattern of the plurality of colors,
It becomes easier to detect color unevenness.

According to an eighth aspect of the present invention, in the color tone inspection method according to the first aspect, in the step of dividing into the plurality of divided areas, the generated tone data image is divided for each predetermined area as a minimum reference. Divided into unit areas and divided into a plurality of minimum areas, a statistical distribution is determined for the color tone data for each of these minimum areas, and each of the plurality of minimum areas has a statistical distribution value in at least one predetermined distribution range. It is characterized in that it is classified into a set that can be included and a set that cannot be included and is divided into the plurality of divided areas. According to this method, it is possible to divide the inspection target into a plurality of divided areas in accordance with the pattern of the plurality of colors, and it is easy to detect color unevenness.

According to a ninth aspect of the present invention, in the color tone inspection method according to any one of the first to eighth aspects, the predetermined inspection reference value is generated when a plurality of non-defective products are used as the inspection object. The statistical calculation is performed on each of the plurality of divided regions for each of the plurality of toned data images, and the statistical calculation is performed on each of the plurality of divided regions. The method is characterized in that each of the plurality of divided areas is individually set based on a variation range of the statistical calculation value. According to this method, it is possible to perform a color tone inspection according to a plurality of non-defective products. In addition, it is very difficult and troublesome to determine a color tone inspection reference value for each divided area of an inspection object having a pattern or to input a numerical value. The system automatically measures and calculates the statistical calculation value of the color tone data for each divided area, and determines the non-defective reference range for each divided area based on this statistical calculation value, so that the variation range of the actually produced non-defective product is It is possible to precisely grasp the inspection reference value based on the ability and to automatically and easily set the inspection reference value.

According to a tenth aspect of the present invention, in the color tone inspection method according to any one of the first to ninth aspects, in the step of performing the statistical calculation, the color tone data forming an occurrence frequency distribution of the color tone data in the divided area is provided. Is obtained as the statistical calculation value, and in the step of performing the comparison, the ratio of the frequency of the predetermined color tone data to the entire inspection object is compared with the predetermined ratio. According to this method, color unevenness inspection can be performed by relatively simple processing.

According to an eleventh aspect of the present invention, in the color tone inspection method according to any one of the first to ninth aspects, in the step of performing the statistical calculation, the color tone data forming an occurrence frequency distribution of the color tone data in the divided area is provided. Is obtained as the statistical calculation value, and the comparing is performed by comparing the shape of the occurrence frequency distribution with the shape of a predetermined reference distribution. According to this method, it is possible to detect the color unevenness feature with high accuracy.

The invention according to claim 12 is the invention according to claims 1 to 11
In the color tone inspection method according to any one of the above, after the step of performing the color tone inspection, a step of displaying the color tone data image on display means for displaying various information on a screen; and a step of displaying the color tone data displayed on the display means. When an operation input for designating a desired divided region through the input means for various operation inputs among the plurality of divided regions in the image is obtained,
Displaying the statistical calculation value or the inspection result of the designated divided area on the display means. In the displaying the color tone data image, the color tone data image is converted into each pixel constituting the color tone data image. A corresponding display color or shading is assigned to the color tone data and displayed on the display means. According to this method,
The display is easy to see and can be compared with visual observation.

According to a thirteenth aspect of the present invention, in the color tone inspection method according to the twelfth aspect, the method further comprises a step of converting a spectral value at each point on the surface into color data corresponding to a visual appearance to generate a real image. In the step of displaying the color tone data image, the real image is displayed together with the color tone data image. According to this method, the display is easy to see, it is possible to make a comparison with the visual inspection, and it is possible to make a precise judgment of the semi-automatic visual inspection.

According to a fourteenth aspect of the present invention, in the color tone inspection method according to the twelfth or thirteenth aspect, at least the step of displaying the color tone data image includes a defect candidate in the comparison result for each of the plurality of divided regions. If the comparison result indicating that the defect region is included, the divided region that has become the comparison result indicating the defect candidate is highlighted. According to this method, it becomes easy to determine a divided region (a portion where there is a possibility of abnormality) of a defect candidate, and it is possible to make a visual determination efficiently and reduce oversight.

The invention according to claim 15 is the invention according to claims 12 to 1.
4. In the color tone inspection method according to any one of (4) and (4), after at least the step of displaying the color tone data image, if a visual pass / fail judgment is input by an operation input through the input means, the visual pass / fail judgment information is displayed. Storing in a predetermined storage means together with information obtained from a spectral spectrum value at each point on the surface and information on a predetermined inspection reference value, and an automatic pass / fail judgment result and a pass / fail judgment obtained from a result of the color tone inspection. Updating the predetermined inspection reference value so that the difference rate is lower than the predetermined rate, using the information stored in the storage unit, when the difference rate indicating the different rate is equal to or more than a predetermined rate; Is further provided. According to this method, the predetermined inspection reference value can be set to a value based on the tendency of the process.

The invention of claim 16 is the invention of claims 12 to 1.
5. In the color tone inspection method according to any one of the aspects 5, in the step of displaying the color tone data image, another color tone data image serving as a color tone inspection reference is further displayed on the display unit. According to this method, the color tone inspection criterion (for example, upper and lower limits) becomes clear, visual judgment becomes easy, and high-precision visual judgment becomes possible.

The invention according to claim 17 is the invention according to claims 1 to 16
In the color tone inspection method according to any one of the above, in the step of generating the color tone data image, among the spectral spectrum values at each point on the surface used when calculating the color tone data of each pixel, the inspection object A weighting coefficient for a spectral spectrum value corresponding to a predetermined spectral spectral wavelength related to a color tone variation factor is changed according to a specific color of the inspection object. According to this method, a change occurring due to a specific color tone can be accurately monitored, and a suitable color tone inspection can be performed.

The invention according to claim 18 is the invention according to claims 13 to 1
7. In the color tone inspection method according to any one of the above 7, in the step of generating the real image, after the color tone of each pixel after the conversion into the color data, a weight for a predetermined color tone selected based on visual sensory evaluation The coefficient is changed according to a specific color of the inspection object. According to this method, a change occurring due to a specific color tone can be accurately monitored, and a suitable color tone inspection can be performed.

[0025] The invention of claim 19 is the invention of claims 1 to 18
In the color tone inspection method according to any one of the above, in the step of dividing into the plurality of divided areas, the arrangement of the pattern in the color tone data image, when dividing the color tone data image into the plurality of divided areas, It is characterized in that necessary alignment is performed. According to this method, an inspection without (or almost no) displacement can be performed.

The invention according to claim 20 is the invention according to claims 1 to 18.
In the color tone inspection method according to any one of the above, in the step of dividing into the plurality of divided regions, a partial pattern in the color tone image data serving as an inspection reference and various patterns in the color tone data image obtained from the inspection object. Perform a normalized correlation operation with the partial pattern, and use the position of the partial pattern in the color tone data image obtained from the inspection object, which is the most matching result among the respective results of the normalized correlation operation. And performing positioning necessary for dividing the color tone data image into the plurality of divided areas. According to this method, highly accurate positioning can be performed.

According to a twenty-first aspect of the present invention, in the color tone inspection method according to the nineteenth or twentieth aspect, in the step of generating the color tone data image, each of the surfaces used in calculating the color tone data of each of the pixels. Among the spectral spectral values at points, a weighting coefficient for a spectral spectral value corresponding to a predetermined spectral spectral wavelength selected in advance for the alignment is changed according to a specific color of the inspection object. I do. According to this method, highly accurate positioning can be performed.

The invention according to claim 22 is the invention according to claims 1-21
In the color tone inspection method according to any one of the above, the inspection object is a plate-like building material. According to this method, a color tone inspection of a plate-like building material becomes possible.

According to a twenty-third aspect of the present invention, there is provided a color tone inspection system for detecting a spectral spectrum value at each point on a surface of an inspection object having a pattern of a plurality of colors, and for inspecting the color tone of the surface of the inspection object. The image processing apparatus is configured to calculate the color tone data of each pixel corresponding to each point on the surface from the spectral value at each point on the surface, and to calculate the color tone data based on the color tone data for each pixel. A color tone data image generation function of generating a data image, a division function of dividing the color tone data image into a plurality of divided areas, and a statistical calculation function of performing a statistical calculation on color tone data of each pixel in each of the plurality of divided areas. A comparison function for comparing the statistical calculation value obtained by the statistical calculation function with a predetermined inspection reference value, and a table of the inspection object according to the comparison result. And having a color tone test function of the color testing.

With this configuration, a color tone data image is generated from the spectral values at each point on the surface of the inspection object, and the color tone data image is divided into a plurality of divided areas.
A statistical calculation is performed on the color tone data of each pixel in each of the plurality of divided regions, a comparison between the statistical calculation value and a predetermined inspection reference value is performed, and the color tone of the surface of the inspection object is determined in accordance with the comparison result. The inspection will be performed. As a result, a high-precision spectral spectrum value can be obtained from the spectroscopic sensor, so that high-precision color tone measurement can be performed. In addition, since the statistical calculation is performed in units of divided regions, there is no problem with visual inspection without being affected by fine patterns. Can be detected stably, and a color tone inspection of an inspection object having a pattern of a plurality of colors becomes possible.

[0031]

FIG. 1 is a block diagram of a color tone inspection system, and FIG. 2 is a diagram for explaining the operation of the entire color tone inspection system of FIG. 1. A first embodiment according to the present invention will be described with reference to these drawings. The form will be described.

As shown in FIGS. 1 and 2, the color tone inspection system according to the first embodiment is capable of imaging 60 times per second and has a spectral spectrum at each point on the surface of the inspection object T having a pattern of a plurality of colors. A spectral camera (spectral sensor) 1 for detecting a value, an input device 2 such as a keyboard and a mouse (not shown), a display device 3 such as a CRT or an LCD, and the spectral camera 1, the input device 2 and the display device 3 are connected. And an image processing device 4 for inspecting the color tone of the surface of the inspection object T.

The image processing device 4 includes a CPU, a semiconductor storage device, a recording device (for example, a hard disk device),
And a program executed by the CPU. The color tone data of each pixel corresponding to each point on the surface is calculated from the spectral value at each point on the surface of the inspection object T, and the color tone data of each pixel is calculated. A tone data image generating function 41 for generating a tone data image, and a dividing function 42 for dividing the tone data image into a plurality of divided areas.
A statistical calculation function 43 for performing a statistical calculation on the color tone data of each pixel in each of the plurality of divided areas; and a comparison function for comparing the statistical calculation value obtained by the statistical calculation function 43 with a predetermined inspection reference value 44, and a color tone inspection function 45 for inspecting the color tone of the surface of the inspection object T according to the comparison result.

FIGS. 3 and 4 further show how the spectroscopic camera of FIG. 2 captures an image of the inspection object. 2 and 3, the inspection object T is a colored plate-like member, for example, an exterior building material, and is conveyed at a speed of 60 m / min in one direction by a conveyor (not shown). Is irradiated with linear light along the lateral direction of the inspection object T by the line light source 5. Here, the line light source 5 is provided with a halogen lamp, and the light from the halogen lamp is adjusted so as to illuminate the field of view almost uniformly by the line type fiber.

The light source of the line light source 5 is not limited to a halogen lamp, but may be a xenon lamp. It is preferable that the light source be a light source that has a uniform wavelength characteristic, has no bright line, is stable, and has little fluctuation. Although a line type fiber is used in this embodiment, a rod-like light source such as a fluorescent lamp may be used. Further, a line light source may be configured by arranging a large number of LEDs (light emitting diodes) in a line. If the emission spectra of the LEDs are not uniform, it is preferable to achieve uniform emission spectra as a line light source by appropriately dispersing and disposing a plurality of types of LEDs having different emission spectra.

The light emitted from the line light source 5 to the surface of the inspection object T is reflected on the surface of the inspection object T,
Light is received by the spectroscopic camera 1 fixed above the surface of the inspection target T so that the entire inspection target T falls within the visual field range. In the examples shown in FIGS. 2 and 4, the spectral camera 1 is provided with a light receiving lens 11, a prism or a diffraction grating spectral filter (prism in FIG. 2) provided behind the lens 11, and provided behind the lens. A two-dimensional CCD element 13;
Since the slit 10 is provided between the lens 11 and the spectral filter 12 in a direction orthogonal to the transport direction of the inspection target T, the light reflected on the surface of the inspection target T passes through the lens 11 and the slit. After passing through 10, the light is separated by the spectral filter 12, and thereafter enters the light receiving surface of the CCD element 13. In this configuration, the output result (shading information) of the CCD element 13 differs depending on the presence or absence of the inspection target T. Therefore, if this phenomenon is used, the position of the inspection target T can be detected from the signal of each captured line. Thus, color information of only the inspection target T for each line can be detected.

In the examples of FIGS. 2 and 4, the CCD 1
3 is arranged such that each pixel in the horizontal direction for 512 pixels corresponds to each position on one line in the short direction on the surface of the inspection object T, and each pixel in the vertical direction for 240 pixels is The wavelength at each pixel position in the direction is separated by the spectral filter 12 into wavelengths of 400 nm to 700 n.
Each wavelength at 240 points within the range of m is detected, and the wavelength intensity (brightness) is detected at a level within the range of 0 to 255. That is, the spectral camera 1 measures the intensity of each spectral wavelength at each point in the line direction of the slit 10 on the surface of the inspection object T,
A spectral image I1 is obtained. Since a spectroscopic camera 1 capable of imaging 60 times per second is used, the spectral image I1 for one line is obtained once every 17 ms, at which the inspection object T advances by about 17 mm. By repeating this imaging, data of the spectral image I1 of the entire surface of the inspection target T is obtained. The resolution in the width direction of each line is determined by the size of the field of view, and the resolution in the transport direction is determined by the video rate (60 frames / second) of the spectroscopic camera 1 and the transport speed of the inspection object T. The width is about 17 mm. It is needless to say that the use of a higher-speed spectroscopic camera makes it possible to further increase the resolution in the transport direction.

FIGS. 5 and 6 show typical examples of spectral images obtained by the above-mentioned spectral camera. As shown in FIG. 5, the spectroscopic camera 1 uses the spectral patterns of black, red, green, blue, white, and black color patterns (the three primary color bands of R, G, and B, the white band, and the black Band) is imaged and imaged to display device 3
, A spectral image as shown in the schematic example of FIG. 6 is displayed. At this time, red is a long wavelength of visible light, blue is a short wavelength, and green is a color having an intensity in the middle, so that each wavelength region is displayed brightly. Also, white is bright and bright over the whole area because all visible light is mixed, and black is dark and dark because the amount of reflected light is very small.

Returning to FIGS. 1 and 2, the spectral images I1 for each line in the short direction are sequentially taken into the image processing device 4 as image signals, and the color tone data image generating function 41 causes the inspection object T to be inspected. From the spectral value at each point on the surface,
A process of calculating color tone data of each pixel corresponding to each point on the surface and generating a color tone data image composed of the color tone data of each pixel is executed.

In the example of FIG. 2, first, the color tone data image generation function 41 (color tone calculation circuit) calculates the integral sum of the intensity of each wavelength and the corresponding predetermined weighting coefficient. The summed data is sequentially stored line by line in an image storage unit constituted by the semiconductor storage element or the recording device. Here, among the spectral spectrum values at each point on the surface of the inspection target T, a weighting coefficient for a spectral spectrum value corresponding to a predetermined spectral spectral wavelength related to a color tone variation factor of the inspection target T is calculated. If the color is changed in accordance with the specific color of T, a change occurring due to the specific color tone can be accurately monitored, and a suitable color tone inspection can be performed.

FIG. 7 shows an example of color tone conversion by the color tone data image generation function of FIG. After storing the integrated data, the color tone data image generation function 41 converts the data stored in the image storage unit into RGB or Lab data.
In order to obtain a color tone measurement value of the color system of (L ^* a ^* b ^* ) to form a color tone data image, a predetermined color tone conversion calculation process is executed. In the first embodiment, when color differences are compared as in the color inspection of a product, Lab color tone conversion processing is executed in consideration of the fact that Lab is closer to human sensation. Thereby, a real image as a color restoration image, and an L image having a color tone measurement value indicated by a Lab display color for each pixel (in FIG. 2, the “inspection object image (L ^*
))), An a image (“inspection object image (a ^* )”), and a b image (“inspection object image (b ^* )”). Since a method of converting the spectral reflectance for each wavelength into Lab is well known, the description thereof is omitted here.

Other typical color systems include Lc
h (L ^* c ^* h ^* ) color system, XYZ color system, Yxy color system, etc. Of these, the Lch color system represents the Lab color system expressed in a polar coordinate system, Like the Lab color system, it is often used for color tone inspection. Where L ^* is brightness,
c ^* is the saturation and h ^* is the hue angle.

When a real image is generated, a weighting coefficient for a predetermined color tone selected based on visual sensory evaluation among color tones of each pixel after conversion to color data is determined by using If the color is changed in accordance with the specific color, a change occurring due to the specific color can be monitored with high accuracy, and a suitable color tone inspection can be performed.

FIG. 8 shows an example of a configuration for obtaining an L image, an a image, and a b image at high speed. In order to generate each of these images, since a spectral image for one line is obtained in units of 17 milliseconds for one frame, it is necessary to process data for a plurality of frames at high speed. For this reason, as shown in the example of FIG. 8, a plurality of CPU boards 401 to 404 that perform processing for a plurality of (four in the same figure) lines continuously taken in the image processing apparatus 4 are provided, respectively. These processing results are taken in via lines L1 to L4, and an L image,
A plurality of CPU boards 4L, 4a, 4b for respectively generating an a image and a b image are provided. In this way, the spectral images of the respective lines are distributed and processed by the CPU boards 401 to 404 in the order in which they are taken, and the processing results are obtained by the CPU boards 4L and 4L.
If the configuration is such that the images are transferred to the memories on 4a and 4b and processed, the L image, the a image, and the b image can be obtained at high speed. Note that the L image, the a image, and the b image can be created by a method in which a high-speed CPU processes the spectral image of the entire inspection target T in a short time while storing it in a large-capacity memory.

Returning to FIGS. 1 and 2, the tone data images of the L image, the a image and the b image created in this manner are divided into a plurality of divided areas by the dividing function. Then, the statistical calculation function 43 performs a statistical calculation on the color tone data of each pixel in each of the plurality of divided regions, and compares the statistical calculation value obtained by the statistical calculation function 43 with a predetermined inspection reference value. The color tone inspection function 45 performs a color tone inspection on the surface of the inspection target T according to the comparison result.

In the example of FIG. 2, each value of the L image is averaged in the transport direction of the inspection object T, and the result is arranged on the horizontal axis of the graph and compared with a predetermined threshold value. In this case, since the L value is higher in a portion where there is a color tone abnormality than in other portions, it is possible to perform a color tone test by determining a defect when the value exceeds a threshold value. In addition, by displaying the real image, the L image, the a image, and the b image on the display device 3, the inspection result can be confirmed. Further, by comparing and displaying the measured Lab image and a previously acquired good Lab image, it is possible to confirm the visibility with higher visibility.

FIGS. 9 and 10 are explanatory diagrams of the sorting function of the image processing apparatus. The second embodiment according to the present invention will be described below with reference to these figures. However, the color tone inspection system of the second embodiment has the same configuration as that of FIG. 1 of the first embodiment, and is characterized by the processing executed by the sorting function 42 of the image processing apparatus 4.

As shown in FIG. 9, the inspection object T is conveyed in one direction, firstly, a concavo-convex pattern is formed on the surface by a roll mold 6, and after drying, the roll coater 7 is dried.
The surface is colored by the. At this time, if the original plate of the inspection object T of the exterior building material has uneven thickness or the roll coater 7 is inclined, the coating at the end of the inspection object T may be blurred, or conversely, excessive paint may be applied. Is attached, etc., and color unevenness is caused in the coloring, resulting in streak-like color defects at the ends.

Therefore, as the above-mentioned feature of the second embodiment,
As shown in FIG. 10, the sorting function 42 of the image processing device 4 converts the two-dimensional tone data image I2 generated by the tone data image generating function 41 into at least one side of FIG.
In the example of 0 (b), a plurality of regions are divided so as to be arranged along the shorter side, and divided into a plurality of divided regions R1 to R9.

When the color tone data image I2 is divided into a plurality of divided regions R1 to R9 in this manner, the above-described color unevenness is likely to occur in the divided regions R1 and R9 at both ends in the short direction. , R9 by the statistical calculation function 43
By performing detailed analysis through, for example, it becomes easy to detect a color defect due to the color unevenness. In other words, with a relatively simple process, it is possible to stably detect a slightly larger color unevenness which is a problem visually, without being influenced by a fine pattern.

By the way, since the pattern of the pattern of the inspection object T in the second embodiment has a regularity of repetition along the one direction, the dividing method of FIG. There is also a division method in which a plurality of regions are divided so as to be arranged along a direction orthogonal to the arrangement direction of symbols having repetition regularity in the color tone data image, and divided into a plurality of divided regions. The color tone data image is divided so that a plurality of regions are arranged in a direction orthogonal to the direction in which the color is continuously colored by the roll coater 7, and divided into a plurality of divided regions. According to the former division method, it is easy to detect a color tone abnormality related to the regularity of the pattern of the lattice pattern, and according to the latter division method, continuous stability such as paint fading of the roll coater 7 is inspected. It will be easier.

The division is not limited to such a division method in the direction, and the generated color tone data image is divided so that a plurality of regions are arranged along a longitudinal side, and divided into a plurality of division regions. It may be a method. According to this division method, slightly larger color unevenness which is a problem visually can be stably detected by relatively simple processing without being influenced by fine patterns.

Further, this dividing method divides the generated tone data image so that a plurality of regions are arranged along a direction in which symbols having repetition regularity in the tone data image are arranged. This is also a division method that divides the generated tone data image so that a plurality of regions are arranged along a direction in which the roll coater 7 successively colors the color data image, and divides the generated tone data image into a plurality of divided regions. It is also a division method. According to the former division method, it is easy to detect a color tone abnormality that is less relevant to the regularity of the pattern of the lattice pattern, and according to the latter division method, the roll coater 7 can be easily detected.
It is easy to check the color balance and color unevenness in the width direction of the coloring, and particularly, it is possible to detect a color defect caused by running out of paint on the roll coater 7 while the front end is normally printed. Become.

FIGS. 11 and 12 are explanatory diagrams of the sorting function of the image processing apparatus. The third embodiment according to the present invention will be described below with reference to these drawings.

The color tone inspection system of the third embodiment is different from the second embodiment in the processing method executed by the sorting function 42 of the image processing apparatus 4. That is, as shown in FIG. 11A, the division function 42 of the third embodiment divides the two-dimensional color tone data image I2 generated by the color tone data image generation function 41 by the division unit area DR serving as the minimum reference. To obtain a statistical distribution of average values of L value, a value, and b value with respect to the tone data of each of the minimum regions, and calculate each of the plurality of minimum regions by a value of the statistical distribution. (Average value) is classified into a group that falls within at least one predetermined distribution range and a group that does not fall in at least one predetermined distribution range, and is divided into a plurality of divided regions.

In the color tone inspection system having this configuration, the color tone data image obtained from the inspection object T having the pattern divided into the joint paint portion and the top paint portion serving as the base is divided by the sorting function 42 into the area of the joint paint portion. It will be divided into two divided areas with the area of the top coating part.

Here, the predetermined distribution range (region determination value in FIG. 11B) is determined by using a plurality of non-defective products as the inspection object T.
L in the minimum area unit shown in each of FIGS.
It is determined in advance based on the frequency distribution of the average value, the value a, and the value b. That is, in the case of a non-defective product, as shown in FIG.
Values, a value, and average value of b value are L3 to L4, a1, respectively.
Ａa2, b3〜b4, and the average value of the L value, a value, and b value of the minimum area unit for the joint coating portion is L1.
When L2, a3 to a4, and b1 to b2,
As shown in (b), the area determination value for the top painted portion is determined from (L3, a1, b3) to (L4, a2, b4), and the area determination value for the joint painted portion is (L1, a3,
b1) to (L2, a4, b2).

If the average of the L value, the a value, and the b value in each of the plurality of divided minimum areas is the area determination value for the top coating portion, the top coating shown in FIG. If the minimum area is included in the joint (Lt, at, bt) and the area determination value for the joint painting part is set, the process of including the minimum area in the joint painting part (Lm, am, bm) is executed. If this processing is executed for all the minimum areas, the color tone data image is divided into the area of the top painting section shown in FIG. 11C and the area of the joint painting section shown in FIG. Can be divided into
In the third embodiment, since the combination is classified into a set that falls within two types of predetermined distribution ranges and a set that does not fall within the predetermined distribution range, FIG.
Each region of (d) does not include the boundary between the top painted portion and the joint painted portion. The same processing as that in the first embodiment is performed on the divided areas divided in this way, but in this case, it is easier to detect color unevenness.

In the third embodiment, a method is employed in which the generated color tone data image is divided into a plurality of minimum regions by dividing the generated color tone data image into division unit regions serving as minimum references. For each predetermined area, a method may be used in which the area is divided into a plurality of minimum areas by dividing by a division unit area serving as a minimum reference. This method also makes it easier to detect color unevenness. However, the predetermined area is divided in the same manner as the division method of the second embodiment or a division method related thereto.

FIG. 13 is an explanatory diagram of a statistical calculation function and a comparison function of the image processing apparatus. A fourth embodiment according to the present invention will be described below with reference to FIG. However, it is assumed that the divided areas R1 to R9 in FIG. 13 correspond to those in FIG. 10B of the second embodiment. Further, each value in FIG. 13 corresponds to any one type of Lab, and the other values are not shown.

The color tone inspection system according to the fourth embodiment includes the first
The configuration is the same as that of FIG. 1 of the embodiment, and is characterized by each processing executed by the statistical calculation function 43 and the comparison function 44 of the image processing apparatus 4. That is, as shown in FIG. 13A, the statistical calculation is performed on the color tone data of each pixel in each of the plurality of divided regions, and as shown in FIG.
The distribution of each average value (b) in FIG.
3 is required.

Then, the comparison function 44 compares the average value of each divided area with a predetermined inspection reference value. Here, the predetermined inspection reference value is derived from a plurality of tone data images generated when a plurality of non-defective products are used as the inspection object T, and is determined for each of the plurality of tone data images. The above-described statistical calculation is performed for each of the plurality of divided regions, and individually set for each of the plurality of divided regions based on the variation range of the statistical calculation value obtained for each of the divided regions. That is, the predetermined inspection reference value is determined in advance based on each of the average values of the divided areas obtained using a plurality of non-defective products as the inspection target T. In the example of FIG. 13A, a white point is a median of a predetermined inspection reference value, and a width defined by an arrow extending vertically from the white point is a management width of the predetermined inspection reference value.

Therefore, the comparison function 44 compares whether or not the measured average value of the divided area unit falls within the management range of the corresponding inspection reference value.
In accordance with the result of the comparison in 4, the color tone inspection of the surface of the inspection object T is performed by the color tone inspection function 45. For example, if the measured average value of each divided area falls within the management range of the corresponding inspection reference value, it is determined that there is no problem such as color unevenness in the divided area, but it must be within the management width. For example, it is determined that the divided area has a problem such as color unevenness. The results of these color tone tests are shown in FIG.
After normalization as shown in (b), it is displayed on the display device 3.

In the fourth embodiment, the average value of each Lab is used for statistical calculation and comparison. However, the present invention is not limited to this. The frequency of the color tone data forming the frequency distribution may be obtained as a statistical calculation value, and the ratio of the frequency of the predetermined color tone data to the entire inspection object T may be compared with the predetermined ratio. For example, as shown in FIG.
A frequency of 70 is obtained, and a ratio (%) of the frequency to the whole inspection object T (%) is compared with a predetermined ratio (70%). If the former is equal to or greater than the latter, the inspection object T is regarded as a good product. Otherwise, it is considered a defective product. In this case, it is possible to inspect color unevenness by relatively simple processing.

Alternatively, the frequency of the color tone data forming the frequency distribution of color tone data in the divided area may be obtained as a statistical calculation value, and the shape of the frequency distribution may be compared with the shape of a predetermined reference distribution. For example, as shown in FIG. 15, the shape of the solid line occurrence frequency distribution is determined, and a comparison is made to determine whether or not this shape is within a range sandwiched between a pair of reference distribution shapes indicated by dotted lines. If the inspection target T is a good product, the test object T is a defective product. In this case, it is possible to detect the color unevenness feature with high accuracy.

Further, in the fourth embodiment, the measured values (L
The average value of each of the ab) may be stored in a storage device or the like, and the average value of the measured values of the divided areas may be obtained and updated to the median of the inspection reference values.

Further, the average value of the whole manufactured inspection object T may be displayed on the display device 3 in chronological order. In this case, color tone change tendency management is possible.

FIG. 16 is a block diagram of the color tone inspection system, and FIG. 17 is a diagram showing a display example of an image displayed on the display device by the image display processing function of FIG. 16. Referring to FIGS. A fifth embodiment according to the invention will be described.

The color tone inspection system of the fifth embodiment is similar to that of FIG.
As shown in FIG. 6, the configuration is the same as that of the first embodiment except that the image processing apparatus 4A further includes an image display processing function 46.

As shown in FIG. 17, the image display processing function 46 assigns a corresponding display color or shading to the color tone data of each pixel constituting the color tone data image.
Processing for displaying on the display device 3 is performed. Here, FIG.
(A), (b), (c), and (d) show a real image, an L image, an a image, and a b image, respectively.

Here, when an abnormal portion such as color unevenness is detected only by the spectral image I1, a portion having a distinctly different color can be recognized, but it is difficult to immediately recognize a subtly different portion. It is also difficult to recognize which part is abnormal only by displaying the numbers.

For this reason, as shown in FIG. 17, a graph (two-dimensional graph) representing the tone value Lab for each position in a stepwise manner like a contour line is created. By displaying each two-dimensional position with shading, abnormal portions can be clearly recognized. For example, from the maximum value and the minimum value of the L value,
The value in the meantime is divided into 10 levels, and the color tendency of the inspection object T can be grasped by shading according to the difference in brightness from white (maximum value) to black (minimum value). In addition, by displaying the color difference instead of the shading, it becomes even easier to understand. Note that the real image is obtained by reproducing the actual color image from the color information data. Also,
The L image, the a image, and the b image can be obtained by changing the display color density in accordance with the level of the color tone data, or by associating various display colors with the level of the color tone data.

As described above, in the fifth embodiment, since various images as shown in FIG. 17 are displayed on the display device 3, the visibility of the Lab information by the worker is improved, and it is possible to compare the Lab information with the naked eye. Become. In addition, since a real image is displayed together with the color tone data image, precise judgment can be made by a semi-automatic visual inspection.

In the fifth embodiment, the display device 3
When an operation input for designating a desired divided region among the plurality of divided regions in the color tone data image displayed on the input device 2 is obtained, a statistical calculation value or an inspection result for the designated divided region is obtained. The function of displaying on the display device 3 may be provided in the image processing device.

FIG. 18 is a block diagram of the color tone inspection system, FIG. 19 is an explanatory diagram of display contents displayed on the display device by the image display processing function of the image processing device, and FIGS.
Is a diagram showing a display example of an image displayed on the display device,
The sixth embodiment according to the present invention will be described below with reference to these drawings.

The color tone inspection system according to the sixth embodiment
As a difference from the embodiment, the processing method executed by the image display processing function 46 is different, and the image processing apparatus 4B further includes an automatic / visual determination comparison / update function 24.

The image display processing function 46 of the sixth embodiment is
As shown in FIG. 19, in addition to displaying another reference image serving as a color tone inspection reference together with the inspection object image on the display device 3,
If the comparison result for each of the plurality of divided regions includes the comparison result indicating that the defect is a candidate, the divided region serving as the comparison result indicating the defect candidate is emphasized using a frame line. indicate. For example, FIG.
20 (a), (b), (c), and (d) together with a real image, an L image, an a image, and a b image as inspection object images shown in (f), (g), and (h), respectively. Real image, L image, a image, b
The image is displayed on the display device 3. Thereby, it is possible to clearly recognize the abnormal point. In addition, the divided area that is a comparison result indicating a defect candidate is highlighted by blinking a frame line surrounding the corresponding part as shown in FIGS. 21 (a) and 21 (b). As a result, the defect candidate portion (the portion having a possibility of abnormality) can be easily identified, efficient visual determination can be performed, and oversight by visual determination can be reduced. Also, real images, L images,
Each of the a-image and the b-image or any reference image may be divided into an upper-limit reference image and a lower-limit reference image, and an image indicating the non-defective product limit may be displayed. As a result, the inspection determination reference range becomes clear, which facilitates determination near the limit, and further improves inspection accuracy.

The automatic / visual judgment comparison / updating function 24, when a visual pass / fail judgment is input by an operation input through the input device 2, transmits the visual pass / fail judgment information to information (for example, color tone data ) And the information of the predetermined inspection reference value, for example, stored in the recording device described above,
The difference rate of the match / mismatch history between the automatic pass / fail judgment result by the color tone inspection function 45 and the input pass / fail judgment is calculated, and the calculated difference ratio is displayed on the display device 3 together with the current match / mismatch result.
Perform the processing to be displayed in. In FIG. 19, the pass / fail judgment by visual observation is “good” or “bad” in “visual judgment input”.
Is input by clicking the radio button of the input device 2 with a mouse or the like. If the pass / fail judgment matches the automatic pass / fail judgment result (the automatic judgment result in the figure) by the color tone inspection function 45, "match" is displayed in a column corresponding to "collation" shown in FIG. If not, "mismatch" is displayed. In addition, the difference rate of the history is "mismatch rate"
Is displayed in% in the column corresponding to. In FIG. 19, the result of the automatic determination is “bad”.

The automatic / visual judgment comparison / updating function 24 uses the information stored in the recording device to set the difference rate to a predetermined rate when the difference rate (that is, the mismatch rate) of the histories exceeds a variable predetermined rate. Further, a process of updating the predetermined inspection reference value so as to fall below the predetermined value is further performed. In the example of FIG. 19, the variable predetermined rate serving as an update reference is set to 1%, and the date when the difference rate exceeds this rate is displayed in the “update history” column together with the value of the determination parameter. I have. Thereby, even if a difference between the visual judgment criterion and the automatic judgment criterion occurs due to a process variation,
It becomes possible to correct the automatic judgment criterion based on the visual judgment. Also, at this time, the past inspection record data is re-inspected and simulated with the newly updated determination parameter to check the difference rate, and the newly updated determination parameter can be adjusted to an appropriate value. Therefore, it is easy to improve the difference rate with high reliability.

According to such a configuration, the color tone inspection relating to the human sensibility is added to the judgment of the quality by visual inspection of the human, so that a preferable inspection result can be obtained. In addition, when a person is tired by a long-time operation, an error is likely to be made in the determination. However, since the inspection object image and the reference image are displayed for comparison, erroneous determination is reduced and work efficiency is improved. Further, when the discrepancy between the automatic pass / fail judgment result and the pass / fail judgment by a person increases, a predetermined test reference value (judgment parameter) is updated based on the stored test information. Even if the automatic pass / fail judgment is not obtained, the repetitive inspection enables the automatic pass / fail judgment based on the tendency of the process. Further, since the change date of the determination parameter and the values before and after the change are displayed, workability is improved. Further, since the divided area which is the comparison result indicating the defect candidate is highlighted, the defective portion can be easily specified.

In each of the above embodiments, when dividing into a plurality of divided areas, the alignment necessary for dividing the toned data image into a plurality of divided areas using the arrangement of symbols in the toned data image is used. May be performed. According to this configuration, the position of the pattern or the printing differs depending on the product in the exterior building material or the like, and a position shift occurs. Even in such a case, an inspection with no (or almost no) position shift can be performed.

Alternatively, a normalization correlation operation is performed between a partial pattern in the color tone image data serving as an inspection reference and various partial patterns in the color tone data image obtained from the inspection object T, and these normalized correlation operations are performed. This is necessary when the color tone data image is divided into a plurality of divided regions by using the position of the partial pattern in the color tone data image obtained from the inspection target T, which is the best matching result among the results. A configuration for performing alignment may be used. According to this configuration, highly accurate positioning can be performed.

When these alignments are performed,
Among the spectral values at each point on the surface used when calculating the color tone data of each pixel, a weighting factor for a spectral value corresponding to a predetermined spectral wavelength selected in advance for alignment is inspected. The configuration may be changed in accordance with the specific color of the object T. In this configuration, it is easy to detect the edge of the pattern for detecting the positional deviation, and it is possible to perform the positioning with higher accuracy.

For example, in the case of poor color tone such as the amount and ratio of a specific paint component, the variation factor is often determined, and when the quality standard for a specific color is strict in detecting a process defect, the variation in the specific factor is easily detected. The modified color tone data (Lab) may be generated by increasing the weighting coefficient for the wavelength of the important spectral spectrum. Specifically, when the paint used has a high reflectance with respect to light having a wavelength of 600 ± 5 nm, the weight coefficient for color tone conversion is set to 60.
The data multiplied by 1.2 at 0 ± 5 nm is defined as color tone data. In this case, when comparing the exterior building material coated with the paint with the exterior building material not coated, the weight coefficient was 600 ±
If the color difference is not 1.2 times at 5 nm, the color difference is 0.2.
Compared to 8, the color difference becomes 3.2 in the case of 1.2 times, so that the characteristics of the paint can be detected in more detail.

In addition, new tone data (Lab) is generated so that the difference between the two tones sandwiching the reference edge becomes remarkable, or weight is added to a specific waveform in which misregistration is easily detected. Is also good. Further, if any one of the color tone data images of Lab is suitable for alignment, the alignment may be performed based on the one color tone data image.

[0086]

As is apparent from the above description, the invention according to claim 1 performs image processing using a spectral sensor that detects spectral spectral values at respective points on the surface of an inspection object having a pattern of a plurality of colors. The apparatus is a color tone inspection method for inspecting the color tone of the surface of the inspection object, from the spectral spectrum value at each point of the surface, to calculate the color tone data of each pixel corresponding to each point of the surface, Generating a tone data image composed of the tone data of each of the pixels, dividing the tone data image into a plurality of divided areas, and performing a statistical calculation on the tone data of each pixel in each of the plurality of divided areas A step of comparing the statistical calculation value obtained in this step with a predetermined inspection reference value; and a color tone of the surface of the inspection object according to the comparison result. Since a step of the 査 enables detailed color measurement, it is possible shades inspection of the inspection object having a pattern of plural colors.

According to a second aspect of the present invention, in the color tone inspection method according to the first aspect, the color tone data image is two-dimensional image data, and in the step of dividing into the plurality of divided regions, The color tone data image is divided so that a plurality of regions are arranged along at least one side, and divided into the plurality of divided regions. It is possible to stably detect a slightly larger color unevenness which is a problem.

According to a third aspect of the present invention, in the color tone inspection method according to the first aspect, the pattern of the pattern of the inspection object has a repetitive regularity, and the color tone data image is two-dimensional image data. In the step of dividing the generated tone data image into a plurality of divided regions, the generated tone data image is arranged such that a plurality of regions are arranged in a direction orthogonal to a direction in which the symbols having the repetition regularity in the tone data image are arranged. And the image is divided into the plurality of divided areas, so that it is easy to detect a color tone abnormality related to, for example, the regularity of the pattern of the lattice pattern.

According to a fourth aspect of the present invention, in the color tone inspection method of the first aspect, the pattern of the pattern of the inspection object has a repetitive regularity, and the color tone data image is two-dimensional image data. In the step of dividing into the plurality of divided regions, the generated tone data image is divided so that a plurality of regions are arranged along a direction in which symbols having the repetition regularity in the tone data image are arranged. Since the image is divided into the plurality of divided areas, it is easy to detect a color tone abnormality that is less relevant to the regularity of the pattern of the lattice pattern, for example.

According to a fifth aspect of the present invention, in the color tone inspection method according to the first aspect, the inspection object is colored continuously from one end to the other by a coloring device, and the color tone data is obtained. The image is two-dimensional image data, and in the step of dividing the image into the plurality of divided regions, the generated color tone data image is divided into a plurality of images along a direction orthogonal to a direction continuously colored by the coloring device. Since the area is divided so as to be arranged side by side and divided into the plurality of divided areas, it is easy to inspect the continuous stability of the coloring apparatus, such as paint faintness.

According to a sixth aspect of the present invention, in the color tone inspection method according to the first aspect, the inspection object is colored continuously from one end to another by a coloring device, and the color tone data The image is two-dimensional image data, and in the step of dividing into the plurality of divided regions, the generated tone data image is arranged such that a plurality of regions are arranged along a direction in which the coloring device continuously colors the image. , And is divided into the plurality of divided areas, so that it is easy to inspect the color balance and the color unevenness in the coloring width direction of the coloring apparatus.

According to a seventh aspect of the present invention, in the color tone inspection method according to the first aspect, in the step of dividing into the plurality of divided areas, the generated tone data image is divided by a division unit area serving as a minimum reference. Then, a statistical distribution is obtained for the color tone data for each of the minimum areas, and each of the plurality of minimum areas is included in a set whose statistical distribution value falls within at least one predetermined distribution range. Since the test object is classified into a plurality of divided areas and divided into the plurality of divided areas, it can be divided into a plurality of divided areas in accordance with a pattern of a plurality of colors of the inspection object, and color unevenness can be easily detected.

According to the eighth aspect of the present invention, in the color tone inspection method according to the first aspect, in the step of dividing into the plurality of divided areas, the generated tone data image is set to a minimum reference value for each predetermined area. Divided into a plurality of minimum areas, and a statistical distribution is obtained for the color tone data for each of the minimum areas. Each of the plurality of minimum areas has at least one predetermined distribution having a statistical distribution value. It is classified into a set that does not belong to the range and a set that does not fit, and is divided into the plurality of divided areas, so that it can be divided into a plurality of divided areas according to a pattern of a plurality of colors of the inspection object, and color unevenness is detected. Easier to do.

According to the ninth aspect of the present invention, the first to the fifth aspects are described.
8. In the color tone inspection method according to any one of 8, the predetermined inspection reference value is derived from a plurality of color tone data images generated when a plurality of non-defective products are used for the inspection object. For each of the plurality of tone data images, the statistical calculation is performed for each of the plurality of divided regions, and based on a variation range of the statistical calculation value obtained for each of the divided regions, the plurality of divided regions Each color is individually set, so color tone inspection can be performed for multiple non-defective products, and the variation range of actually produced non-defective products can be accurately grasped in accordance with the process capability. Inspection reference values can be set easily and easily.

According to the tenth aspect, the first aspect is provided.
In the color tone inspection method according to any one of Items 1 to 9, in the step of performing the statistical calculation, a frequency of color tone data forming a frequency distribution of color tone data in the divided region is obtained as the statistical calculation value, and the comparison is performed. In the step, the ratio of the frequency of the predetermined color tone data to the whole inspection object is compared with the predetermined ratio, so that the color unevenness inspection can be performed by a relatively simple process.

According to the eleventh aspect, according to the first aspect,
In the color tone inspection method according to any one of Items 1 to 9, in the step of performing the statistical calculation, a frequency of color tone data forming a frequency distribution of color tone data in the divided region is obtained as the statistical calculation value, and the comparison is performed. In the step, since the shape of the occurrence frequency distribution is compared with the shape of a predetermined reference distribution, the feature of the color unevenness can be detected with high accuracy.

According to the twelfth aspect of the present invention, the first aspect is provided.
12. The color tone inspection method according to any one of claims 11 to 11, further comprising, after the color tone inspection step, displaying the color tone data image on display means for displaying various information on a screen; and displaying the information on the display means. When an operation input for designating a desired divided region among the plurality of divided regions in the color tone data image through input means for various operation inputs is obtained, a statistical calculation value or an inspection result for the designated divided region is obtained. And displaying the color tone data image. Since the information is displayed on the display means, the display is easy to see and can be compared with the visual observation.

According to the thirteenth aspect of the present invention, the first aspect
2. The color tone inspection method according to 2, further comprising the step of converting a spectral spectrum value at each point on the surface into color data corresponding to a visual appearance to generate a real image,
In the step of displaying the color tone data image, the real image is displayed together with the color tone data image, so that the display is easy to see and can be compared with the visual observation, and the precise judgment of the semi-automatic visual inspection can be performed.

According to the fourteenth aspect of the present invention, the first aspect
In the color tone inspection method according to 2 or 13, in the step of displaying at least the color tone data image, if the comparison result for each of the plurality of divided regions includes a comparison result indicating a defect candidate, Since the divided area which is the comparison result indicating the defect candidate is highlighted, the divided area of the defect candidate (the part having a possibility of abnormality) can be easily determined, the visual judgment can be efficiently performed, and the oversight is reduced. .

According to the fifteenth aspect, according to the first aspect,
In the color tone inspection method according to any one of 2 to 14, after at least the step of displaying the color tone data image,
When a visual pass / fail judgment is input by an operation input through the input means, information on the visual pass / fail judgment is given along with information obtained from a spectral spectrum value at each point on the surface and information on a predetermined inspection reference value. The step of storing the information stored in the storage unit when the difference ratio indicating the ratio of the difference between the automatic quality determination result obtained from the result of the color tone inspection and the quality determination becomes a predetermined ratio or more. And updating the predetermined inspection reference value so that the difference rate is lower than the predetermined rate, thereby enabling the predetermined inspection reference value to be set to a value based on the tendency of the process.

According to the sixteenth aspect of the present invention, the first aspect
In the color tone inspection method according to any one of 2 to 15, in the step of displaying the color tone data image, another color tone data image serving as a color tone inspection standard is further displayed on the display means. The lower limit becomes clear, visual judgment becomes easy, and high-precision visual judgment becomes possible.

According to the seventeenth aspect of the present invention, the first aspect
In the color tone inspection method according to any one of Items 1 to 16, in the step of generating the color tone data image, among the spectral spectrum values at each surface point used when calculating the color tone data of each pixel, the inspection is performed. Since the weighting coefficient for the spectral spectrum value corresponding to a predetermined spectral spectral wavelength related to the variation factor of the color tone of the object is changed in accordance with the specific color of the inspection object, the variation generated due to the specific color tone Can be accurately monitored, and a suitable color tone inspection can be performed.

According to the eighteenth aspect, according to the first aspect,
In the color tone inspection method according to any one of Items 3 to 17, in the step of generating the real image, after the color tone of each pixel after conversion to the color data, a predetermined color tone selected based on a visual sensory evaluation Is changed in accordance with the specific color of the inspection object, fluctuations caused by the specific color tone can be monitored with high accuracy, and suitable color tone inspection can be performed.

According to the invention of claim 19, claim 1
18. In the color tone inspection method according to any one of Items 18 to 18, in the step of dividing into the plurality of divided areas, the color tone data image is divided into the plurality of divided areas using an arrangement of symbols in the color tone data image. Since the necessary alignment is performed in such a case, an inspection with no (or little) displacement can be performed.

According to the twentieth aspect, according to the first aspect,
18. In the color tone inspection method according to any one of items 18 to 18, in the step of dividing into the plurality of divided regions, the partial pattern in the color tone image data serving as an inspection reference and the color tone data image in the color tone data image obtained from the inspection target are included. A normalization correlation operation with various partial patterns is performed, and the position of the partial pattern in the color tone data image obtained from the inspection object, which is the most matching result among the results of these normalization correlation operations, is used. Then, since the necessary alignment is performed when the color tone data image is divided into the plurality of divided areas,
High-accuracy alignment is possible.

According to the twenty-first aspect, the first aspect is provided.
21. In the color tone inspection method according to 9 or 20, in the step of generating the color tone data image, of the spectral spectrum values at each point on the surface used when calculating the color tone data of each pixel, The weighting coefficient for the spectral spectrum value corresponding to the predetermined spectral spectral wavelength selected in advance is changed according to the specific color of the inspection object, so that highly accurate positioning can be performed.

According to the twenty-second aspect, the first aspect is provided.
In the color tone inspection method according to any one of Items 21 to 21, since the inspection object is a plate-like building material, the color tone inspection of the plate-like building material becomes possible.

According to a twenty-third aspect of the present invention, there is provided a color tone inspection system for detecting a spectral value at each point on the surface of an inspection object having a pattern of a plurality of colors, and for inspecting the color tone of the surface of the inspection object. The image processing apparatus is configured to calculate the color tone data of each pixel corresponding to each point on the surface from the spectral value at each point on the surface, and to calculate the color tone data based on the color tone data for each pixel. A color tone data image generation function of generating a data image, a division function of dividing the color tone data image into a plurality of divided areas, and a statistical calculation function of performing a statistical calculation on color tone data of each pixel in each of the plurality of divided areas. A comparison function for comparing the statistical calculation value obtained by the statistical calculation function with a predetermined inspection reference value, and a table of the inspection object according to the comparison result. Because it has a color tone test function of the color test enables detailed color measurement, it is possible shades inspection of the inspection object having a pattern of plural colors.

[Brief description of the drawings]

FIG. 1 is a block diagram of a color tone inspection system.

FIG. 2 is an operation explanatory diagram of the entire color tone inspection system of FIG. 1;

FIG. 3 is a diagram showing a state of imaging of an inspection object by the spectral camera of FIG.

FIG. 4 is a diagram showing a state of imaging of an inspection object by the spectral camera of FIG. 2;

FIG. 5 is a diagram illustrating a schematic example of a spectral image or the like obtained by imaging with the spectral camera of FIG. 2;

6 is a diagram illustrating a schematic example of a spectral image obtained by imaging with the spectral camera of FIG. 2;

FIG. 7 is a diagram illustrating an example of color tone conversion by the color tone data image generation function of FIG. 1;

FIG. 8 is a diagram illustrating a configuration example for obtaining an L image, an a image, and a b image at high speed.

FIG. 9 is an explanatory diagram of a sorting function of the image processing apparatus.

FIG. 10 is an explanatory diagram of a sorting function of the image processing apparatus.

FIG. 11 is an explanatory diagram of a sorting function of the image processing apparatus.

FIG. 12 is an explanatory diagram of a sorting function of the image processing apparatus.

FIG. 13 is an explanatory diagram of a statistical calculation function and a comparison function of the image processing apparatus.

FIG. 14 is an explanatory diagram of a statistical calculation function and a comparison function of the image processing apparatus.

FIG. 15 is an explanatory diagram of a statistical calculation function and a comparison function of the image processing apparatus.

FIG. 16 is a block diagram of a color tone inspection system.

17 is a diagram showing a display example of an image displayed on the display device by the image display processing function of FIG. 16, which is a drawing obtained by printing out an intermediate image displayed on the display device by a printer. It is a photograph.

FIG. 18 is a block diagram of a color tone inspection system.

FIG. 19 is an explanatory diagram of display contents displayed on the display device by the image display processing function of the image processing device.

FIG. 20 is a view showing a display example of an image displayed on the display device, and is a photograph replacing a drawing obtained by printing out an intermediate image displayed on the display device by a printer.

FIG. 21 is a diagram illustrating a display example of an image displayed on the display device, and is a photograph replacing a drawing obtained by printing out an intermediate image displayed on the display device by a printer.

[Explanation of symbols]

 Reference Signs List 1 spectral camera 2 input device 3 display device 4 image processing device 41 color tone data image generation function 42 classification function 43 statistical calculation function 44 comparison function 45 color tone inspection function 46 display processing function 47 automatic / visual judgment comparison / update function

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Hidekazu Araki 1048 Kadoma Kadoma, Osaka Prefecture Matsushita Electric Works Co., Ltd. Terms (reference) 2G020 AA04 AA08 CB26 CB32 CB43 CC13 CC42 CD04 CD24 DA05 DA12 DA34 DA42 2G051 AA32 AB11 AB12 BA20 BB17 CA04 CB01 CC15 DA06 EA16 EA17 EC03 5B057 AA12 BA02 CA01 CA08 CA12 CA16 CE09 DC06 DB09 A09 BA03 CA02 DA02 FA37 FA45 GA19 GA41 JA11

Claims (23)

[Claims] An image processing apparatus uses a spectral sensor that detects a spectral spectrum value at each point on a surface of an inspection target having a pattern of a plurality of colors to perform a color tone inspection for a color tone of the surface of the inspection target. In the inspection method, a step of calculating color tone data of each pixel corresponding to each surface point from the spectral spectrum value at each surface point, and generating a color tone data image including the color tone data of each pixel, A step of dividing the color tone data image into a plurality of divided areas; a step of performing a statistical calculation on the color tone data of each pixel in each of the plurality of divided areas; a statistical calculation value obtained in this step; Comparing with a value, and performing a color tone inspection of the surface of the inspection object according to the comparison result. Method. 2. The method according to claim 1, wherein the tone data image is two-dimensional image data.
2. The color tone inspection method according to claim 1, wherein a plurality of regions are divided along the side so as to be divided into the plurality of divided regions. 3. The pattern of the pattern of the inspection object has a repetitive regularity, the color tone data image is two-dimensional image data, and in the step of dividing into the plurality of divided areas, the generated color tone data The image is divided so that a plurality of regions are arranged along a direction orthogonal to the arrangement direction of the symbols having the repetition regularity in the color tone data image, and the image is divided into the plurality of divided regions. The color tone inspection method according to claim 1. 4. The arrangement of the pattern of the inspection object has a repetitive regularity, the color tone data image is two-dimensional image data, and in the step of dividing into the plurality of divided areas, the generated color tone data 2. The image is divided into a plurality of regions along a direction in which the symbols having the repetition regularity in the color tone data image are arranged, and the image is divided into the plurality of divided regions. 3. Color tone inspection method. 5. The inspection object is colored continuously from one end to another by a coloring device, and the color tone data image is two-dimensional image data, and is divided into the plurality of divided areas. In the step, the generated tone data image is divided into a plurality of regions along a direction orthogonal to a direction in which the color is continuously colored by the coloring device, and divided into the plurality of divided regions. The color tone inspection method according to claim 1, wherein: 6. The inspection object is colored continuously from end to end by a coloring device, and the color tone data image is two-dimensional image data, and is divided into the plurality of divided areas. In the step, the generated tone data image is divided so that a plurality of regions are arranged along a direction colored continuously by the coloring device, and is divided into the plurality of divided regions. The color tone inspection method according to claim 1. 7. In the step of dividing the color tone data image into a plurality of minimum regions, the generated tone data image is divided into a plurality of minimum regions by dividing the generated tone data image into a plurality of minimum regions. Calculating a statistical distribution for each of the plurality of minimum regions, classifying each of the plurality of minimum regions into a set and a set not including a value of the statistical distribution within at least one predetermined distribution range, and dividing the plurality of divided regions into the plurality of divided regions. The color tone inspection method according to claim 1, wherein: 8. In the step of dividing the color tone data image into a plurality of divided areas, the generated color tone data image is
It divides into a plurality of minimum areas by dividing by a division unit area serving as a minimum reference, obtains a statistical distribution for color tone data for each of the minimum areas, and calculates a value of the statistical distribution of each of the plurality of minimum areas by at least one. 2. The color tone inspection method according to claim 1, wherein the plurality of divided areas are classified into sets that fall within two predetermined distribution ranges and sets that do not. 9. The predetermined inspection reference value is derived from a plurality of tone data images generated when a plurality of non-defective products are used as the inspection target, and each of the plurality of tone data images is With respect to each of the plurality of divided regions, the statistical calculation is performed for each of the plurality of divided regions, and a variation range of the statistical calculation value of each of the plurality of divided regions obtained as a reference is individually set for each of the plurality of divided regions. 9. The color tone inspection method according to claim 1, wherein the color tone inspection method is set. 10. The step of performing the statistical calculation obtains, as the statistical calculation value, the frequency of the color tone data that forms the frequency distribution of the color tone data in the divided region, and the step of performing the comparison includes the step of: The color tone inspection method according to any one of claims 1 to 9, wherein a ratio of a frequency to the whole inspection object is compared with a predetermined ratio. 11. The step of performing the statistical calculation obtains a frequency of color tone data forming a frequency distribution of color tone data in the divided area as the statistical calculation value, and the step of performing the comparison includes the step of calculating the frequency of the frequency distribution of the occurrence. The shape is compared with a shape of a predetermined reference distribution.
9. The color tone inspection method according to any one of 9. 12. A step of displaying the color data image on display means for displaying various information on a screen after the step of performing the color test, and the plurality of divisions in the color data image displayed on the display means. Displaying, on the display unit, a statistical calculation value or an inspection result of the specified divided region when an operation input for designating a desired divided region out of the regions through the input means for various operation inputs is obtained; In the step of displaying the color tone data image, the color tone data image is converted into color tone data of each pixel constituting the color tone data image,
The color tone inspection method according to claim 1, wherein a corresponding display color or shade is assigned and displayed on the display unit. 13. The method according to claim 13, further comprising: converting a spectral value at each point on the surface into color data corresponding to a visual appearance to generate a real image. 13. The color tone inspection method according to claim 12, wherein the real image is displayed together with the real image. 14. At least at the step of displaying the color tone data image, if the comparison result for each of the plurality of divided regions includes a comparison result indicating a defect candidate, the image is determined to be the defect candidate. 14. The color tone inspection method according to claim 12, wherein a divided area which is a comparison result indicating a color difference is highlighted. 15. At least after the step of displaying the color tone data image, when a visual pass / fail judgment is input by an operation input through the input means, the visual pass / fail judgment information is analyzed by spectral analysis at each point on the surface. A step of storing the information obtained from the spectrum value and the information of the predetermined inspection reference value in a predetermined storage means, and a difference rate indicating a ratio of a difference between the automatic quality judgment result obtained from the color tone inspection result and the quality judgment. Updating the predetermined inspection reference value such that the difference rate is lower than the predetermined rate, using the information stored in the storage means when the predetermined rate is equal to or more than the predetermined rate. The color tone inspection method according to any one of claims 12 to 14, wherein: 16. The display device according to claim 12, wherein, in the step of displaying the color tone data image, another color tone data image serving as a color tone inspection reference is further displayed on the display means.
16. The color tone inspection method according to any one of items 15 to 15. 17. In the step of generating the color tone data image, among the spectral spectrum values at each point on the surface used when calculating the color tone data of each pixel, the color tone variation factor of the inspection object is determined. 17. The color tone inspection method according to claim 1, wherein a weighting coefficient for a spectrum value corresponding to a related predetermined spectrum wavelength is changed in accordance with a specific color of the inspection object. 18. The method according to claim 18, wherein in the step of generating the real image, after the color tone of each pixel after the conversion into the color data,
18. The color tone inspection method according to claim 13, wherein a weighting factor for a predetermined color tone selected based on visual sensory evaluation is changed according to a specific color of the inspection object. 19. In the step of dividing the color tone data image into a plurality of divided regions, the alignment necessary for dividing the color tone data image into the plurality of divided regions is performed by using a sequence of symbols in the color tone data image. The color tone inspection method according to claim 1, wherein the method is performed. 20. In the step of dividing into a plurality of divided areas, normalization of a partial pattern in color tone image data serving as an inspection reference and various partial patterns in a color tone data image obtained from the inspection object. Performing a correlation operation, using the position of the partial pattern in the tone data image obtained from the inspection object, which results in the best matching among the results of these normalized correlation operations, 19. The color tone inspection method according to claim 1, wherein a necessary alignment is performed when the image is divided into the plurality of divided areas. 21. In the step of generating the color tone data image, among the spectral spectrum values at each point on the surface used in calculating the color tone data of each pixel, the spectral tone values are selected in advance for the alignment. 21. The color tone inspection method according to claim 19, wherein a weighting coefficient for a spectral spectrum value corresponding to a predetermined spectral spectral wavelength is changed according to a specific color of the inspection object. 22. The color tone inspection method according to claim 1, wherein the inspection object is a plate-like building material. 23. An image processing apparatus, comprising: a spectroscopic sensor for detecting a spectral spectrum value at each point on the surface of the inspection object having a pattern of a plurality of colors; and an image processing device for inspecting a color tone of the surface of the inspection object. A processing unit configured to calculate color tone data of each pixel corresponding to each point on the surface from the spectral spectrum value at each point on the surface, and to generate a color tone data image including color tone data of each pixel; A division function for dividing the color tone data image into a plurality of divided areas, a statistical calculation function for performing a statistical calculation on the color tone data of each pixel in each of the plurality of divided areas, and a statistic obtained by the statistical calculation function. It has a comparison function for comparing the calculated value with a predetermined inspection reference value, and a color tone inspection function for inspecting the color tone of the surface of the inspection object according to the comparison result. Color inspection system, characterized in that.