JPH0989762A - Method and apparatus for discrimination of linear material - Google Patents

Abstract

PROBLEM TO BE SOLVED: To judge the wire color and the wire diameter of a dummy wire and to automatically discriminate the kind of the dummy wire. SOLUTION: Total reflecting plane mirrors 11a, 11b are formed at an angled reflecting mirror 11, and a reference plate 12 in which a back reference line 12a is marked on a white ground is arranged and installed. A dummy wire A is set on an inspection part 1. The inspection part 1 is illuminated with an illumination part 2. The reference plate 12, the total reflecting plane mirrors 11a, 11b and the dummy wire A are imaged by a CCD camera head 3. A position in which the reference line 12a is interrupted by the dummy wire is detected so as to find a wire diameter. R-data, G-data and B-data which are obtained by a camera body 4 are converted, by a personal computer 5, into color data which are composed of Y (luminance) data, S (chromaticness) data and H (hue) data. Information on the pixel distribution of the color data as the fundamental color of a wire color is registered in the personal computer 5. The measured color data are collated with the registered pixle distribution, and the wire color is judged.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a linear material discriminating method and apparatus for discriminating the type of a linear material such as an electric wire depending on the color or diameter of the linear material.

[0002]

2. Description of the Related Art Conventionally, in the wire harness manufacturing process,
A process of crimping a predetermined terminal to an electric wire having a predetermined wire color and wire diameter according to the standard is performed by using an automatic cutting and crimping machine. This automatic cutting and crimping machine manufactures various standard products, and when changing the type of manufactured product, it is necessary to inspect whether or not the product according to the standard is manufactured by the automatic cutting and crimping machine. At this time, a dummy wire (test hitting wire) having a length of about 140 mm is prototyped by an automatic cutting and crimping machine, and the dummy wire is inspected so that the switching work of the automatic cutting and crimping machine can be performed smoothly.

[0003]

By the way, the inspection of whether the wire color, wire diameter, and terminal of the dummy wire are manufactured according to the standard has been visually performed manually. This work has a problem that it takes time and labor, especially when the wire diameter is small. For example, the line color may be a combination of two colors in which ten or more basic colors are used and a thin line (stripe) of another color is put in the base of one color, or a single color that is plain. In addition, there are wire diameters of about 1 mm, and it is particularly difficult to visually judge the line color pattern and the wire diameter.

The present invention provides a method and apparatus for discriminating a linear material such as an electric wire, which is capable of automatically discriminating the type of the linear material depending on the color or diameter of the wire. With the goal.

[0005]

The linear material discriminating method of the present invention made to solve the above-mentioned problems is a linear material having a single color among a plurality of predetermined basic colors. Material, or a linear material having a line color that constitutes a stripe in the axial direction with a monochromatic line color and a plurality of basic colors, or a linear material having a line color that constitutes the above stripe,
A method for discriminating a linear material for discriminating a line color of the linear material to discriminate an arbitrary type of the linear material, wherein a surface of the linear material is imaged, and R of a pixel in the imaged image, The G and B data is converted into color data to obtain color data for pixels in a direction orthogonal to the axis of the linear material, the color data is collated with the pixel distribution in the color data for the basic color, and the image is taken. It is characterized in that the line color of the linear material is determined.

Further, the linear material discriminating apparatus of the present invention uses a linear material having a single line color among a plurality of predetermined basic colors or a single color and a plurality of basic colors. Arbitrary linear material by judging the line color of the linear material having a line color that forms a stripe in the axial direction or the linear material that has a line color that forms the stripe Of a linear material for discriminating the type of the linear material, wherein the surface of the linear material is imaged, and R, G, and B of pixels in the imaged image are captured.
Image pickup means for obtaining data, storage means for storing the pixel distribution in the color data for the basic color, and R, G, B data for converting the color data to color data in a direction orthogonal to the axis of the linear material. A line color determining unit that determines color data of pixels, compares the color data with a pixel distribution stored in the storage unit, and determines the line color of the imaged linear member. Characterize.

Further, the linear material discriminating apparatus of the present invention uses a linear material having a single line color among a plurality of predetermined basic colors, or a single color and a plurality of basic colors. Arbitrary linear material by judging the line color of the linear material having a line color that forms a stripe in the axial direction or the linear material that has a line color that forms the stripe A linear material determination device for determining the type of the linear material, the reference line being arranged on the background of the linear material so as to be orthogonal to the linear material, and the reference line and the surface of the linear material. Image pickup means for obtaining R, G, B data of pixels in the picked-up image, storage means for storing the pixel distribution in the color data for the basic color, and the line in the image picked up by the image pickup means. For both sides of the sheet, in the direction intersecting the reference line in the image The element is searched to detect the reference line from the luminance data of the pixel, the pixel is searched along the detected reference line, and the intersection of the reference line and the linear member is obtained from the luminance data of the pixel. A wire diameter determining means for determining the wire diameter of the linear material based on the positions of the intersections on both sides of the linear material;
The G and B data is converted into color data to obtain color data for pixels in the direction orthogonal to the axis of the linear material, and the color data and the pixel distribution stored in the storage unit are collated to capture the image. And a line color determining means for determining the line color of the linear material.

[0008]

According to the method and apparatus for discriminating a linear material of the present invention, the surface of the linear material is imaged, and the R, G, B data of the pixels in the imaged image is converted into color data, and Color data is obtained for pixels in a direction orthogonal to the axis of the material. Then, the obtained color data and the pixel distribution in the color data for the basic color are collated to determine the type of the imaged linear material.

Further, according to the linear material discriminating apparatus of the present invention having the reference line, pixels are first searched in a direction intersecting with the reference line on both sides of the linear material in the image picked up by the image pickup means. Then, the reference line is detected from the luminance data, pixels are searched along this reference line, the intersection point of the reference line and the linear material is obtained from the luminance data, and the positions of the intersection points on both sides of this linear material are obtained. The wire diameter of the linear material is determined based on the. Then, the type of the linear material is discriminated based on the collation of the color data and the pixel distribution and the discrimination result of the wire diameter.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 14 is a diagram showing an example of a dummy line targeted by the determination device of the embodiment of the present invention. The dummy wire is obtained by crimping terminals 20a and 20b on both ends of the electric wire 10 in which the conductor 10a is covered with the insulating material 10b. There are many kinds of dummy wires according to each combination of wire color, wire diameter, and terminal type as shown in Table 1 below.

[0011]

[Table 1]

The term "line color" refers to the types of the single-color pattern and the two-color pattern of the electric wire 10. When referring to the 13 colors of Table 1 that compose this line color, they will be referred to hereafter. It is called "basic color". Further, in Table 1, the wire diameter is shown by the cross-sectional area of the conductive wire 10a, but the actual wire diameter is 2 × √ (S / π) where S is the value in Table 1.

FIG. 1 is a diagram showing a main part of a dummy line discriminating apparatus to which the present invention is applied. This discriminator uses the dummy line A
The inspection unit 1, which sets the inspection unit 1, the illumination unit 2 which irradiates the inspection unit 1 with illumination light, the camera head 3 which captures an image of the dummy line A, the camera body 4 which processes the output signal of the camera head 3, and the type of the dummy line A. Personal computer 5 for processing
And a camera adapter 6 for supplying power to the camera head 3. The inspection unit 1, the illumination unit 2, and the camera head 3 are fixed to members (not shown).

FIG. 2 is a perspective view of the inspection unit 1.
Is a two-sided total reflection plane mirror 11a made of aluminum vapor deposition,
An angled reflection mirror 11 formed with 11b, a reference plate 12 having a black reference line 12a printed on a white background, and a black background plate 13 serving as a background of the terminal portions at both ends of the dummy line A are provided. ing. The total reflection plane mirrors 11a and 11b of the angled reflection mirror 11 form an angle of 120 degrees and meet at the valley line V.
The valley line V is arranged so as to be perpendicular to the optical axis Z of the camera head 3. Further, the reference plate 12 is arranged so that its reference line 12a intersects the valley line V and the optical axis Z at right angles.

In the above structure, the dummy line A is set to the optical axis Z.
The dummy line A is imaged by the camera head 3 by arranging the dummy line A so as to be orthogonal to, and parallel to the valley line V of the total reflection plane mirrors 11a and 11b. At this time, the reference plate 12 and the reference line 12a are located behind the dummy line A, and the back side of the dummy line A is projected at two angles by the total reflection plane mirrors 11a and 11b.

FIG. 3 is a perspective view showing the illuminating section 2. The illuminating section 2 is a dome type diffuser plate 21 having a milky white acrylic plate and a semicircular cross section, and two circular high frequency fluorescent lamps 22a, 22.
a ring illumination 22 in which b is arranged concentrically. The ring illumination 22 is arranged close to the dome type diffusion plate 21 from the side opposite to the inspection unit 1. The illumination unit 2 diffuses the white light of the ring illumination 22 with the dome type diffusion plate 21. The inspection unit 1 is illuminated uniformly. A circular imaging window 21a is formed at the center of the dome-shaped diffusion plate 21, and the camera head 3 captures an image of the inspection unit 1 side through the central space of the ring illumination 22 and the imaging window 21a. It

The camera head 3 is a single plate type RGB color CC.
The D camera outputs R, G, and B luminance signals of each pixel to the camera body 4 with a resolution of 512 × 512 pixels. The camera head 3 is set so that the vertical array direction of the pixels is optically parallel to the reference line 12a in the inspection unit 1. The camera body 4 A / D-converts the R, G, B luminance signals of each pixel and outputs R, G, B data (digital data) which are numerical data of 0 to 255 levels. That is, the R, G, B data is data in which the brightness of each of R, G, B in each pixel is represented by 256 gradations. Then, the personal computer 5 reads the R, G, B data corresponding to each pixel, and determines the type of the dummy line by determining the line diameter, line color and terminal of the dummy line, as described later.

FIG. 4 is a diagram showing an example of a captured image of a dummy line in the embodiment. In the figure, the thickness of the dummy line is exaggerated. Further, in the figure and the following description, unless there is a possibility of confusion, when referring to the image of each member, the same name as the corresponding member is used, and the reference numeral of the corresponding member indicates that the target is an image. Is used with a symbol ''.

FIGS. 4A and 4C show the left and right terminals 20a, 20.
4B is an extracted image corresponding to the central portion of the electric wire 10, and these extracted images are extracted from an image of 512 × 512 pixels corresponding to the angle of view of the camera head 3. It was done. Since the positional relationship between the inspection unit 1 and the camera head 3 is fixed, such a partially extracted image can be easily extracted by deciding the pixel position in advance.

In the extracted image shown in FIG. 4B, a reference plate 12 'and its reference line 12 are provided on one side of the background of the electric wire 10'.
a ′ exists, and next to it, a total reflection plane mirror 11a ′,
11b 'and the electric wire 10' (rear surface) reflected on it are present on both upper and lower sides. Since the illumination light from the illumination unit 2 is white light, the reference plate 12 is a white background, and the total reflection flat mirrors 11a and 11b are mirror surfaces, the reference line 12 'and the electric wire 10' are white. Become. In the extracted images shown in FIGS. 4A and 4C, the left and right terminals 20a ', 20
A black background plate 13 'is present around b'.

The personal computer 5 reads the R, G, B data of the above image for one frame and stores it in the image memory, and the coordinates (position information) of the pixel corresponding to the R, G, B data are stored in the image memory. Based on the determination of the wire diameter,
The line color and terminals are determined.

(Determination of Wire Diameter) FIG. 5 is a diagram for explaining an inspection flow for determining the wire diameter. First, as shown by arrows in FIG. 5A, pixels are sequentially searched in the horizontal direction from a predetermined pixel toward the reference line 12a ', and R, G, and
The B data is read, and it is determined whether or not the luminance data (brightness) of the pixel obtained from the R, G, B data has become equal to or less than a preset threshold value TH1, and the reference line 12a 'is determined.
Position (horizontal coordinate of pixel) is detected. That is, when the luminance data of pixels in the horizontal direction is obtained, this luminance data becomes lower than the threshold value TH1 in the pixel corresponding to the reference line 12a 'as shown in FIG. From this, the position of the reference line 12a 'is detected.

When the position of the reference line 12a 'is detected, as shown in FIG.
As indicated by the arrow (A), the luminance data of the pixel is searched while tracing the reference line 12a 'downward to detect the upper end position (vertical coordinate) of the electric wire 10' and store the value of the vertical coordinate.
That is, the luminance data of the pixel corresponding to the electric wire 10 'as shown in FIG. 5C is lower than the luminance data of the background portion of the reference plate 12' (FIG. 5B) (in the case of dark electric wire color), so that it will be described later. As will be described, the upper end position of the electric wire 10 'is detected from the difference in the brightness data.

Next, as shown by the arrow in FIG. 5A, the luminance data of the pixel is searched from the predetermined pixel in the direction of the reference line 12a 'to detect the position of the reference line 12a'. The brightness data of the pixel is searched while tracing the reference line 12a 'upward as shown by the arrow to detect the lower end position of the electric wire 10' and store the value of the vertical coordinate.

FIG. 6 is a diagram for explaining the logic for detecting the upper end position of the electric wire 10 '. In the following description, the arrangement of pixels in the horizontal direction of the image will be referred to as a “line”. Figure 6
As shown in (A), a detection width of a fixed number of pixels is provided horizontally from the reference line 12a ', and the luminance data of pixels included in the detection width is searched line by line along the reference line 12a'.
At this time, the difference (absolute value) between the luminance data of every two lines is sequentially obtained. Then, when this difference becomes equal to or greater than a preset threshold value TH2, the vertical coordinate of the line between the two lines is determined as the upper end position of the electric wire 10 '.

That is, (i-2) and (i- shown in FIG. 6 (A).
The luminance data in each line of 1), (i), and (i + 1) is shown in FIG.
As shown in (B), (C), (D), and (E), the luminance data in the electric wire 10 'portion becomes lower than the luminance data in the background portion.
Therefore, the difference of the luminance data of each line is calculated for the inspection width including only the reference line 12a 'without the electric wire 10', and the maximum value x? Is adopted as the threshold value TH2. Note that α is an integer greater than or equal to +2 and is a threshold value TH2.
Is, for example, the luminance data in FIG. 6 (E) and FIGS. 6 (B), (C)
, (D) so that the difference with the luminance data is not exceeded. The threshold value TH2 is set in this way, and 2
The difference between the luminance data of lines (i−1) and (i + 1) is the threshold value T.
When it becomes H2 or more, it is determined that the electric wire 10 'is detected,
The vertical coordinate of the line (i) between the two lines (i-1) and (i + 1) is determined as the upper end position of the electric wire 10 '. The lower end position of the electric wire 10 'is also detected by the same logic.

Then, the wire diameter is obtained from the difference (absolute value) between the vertical coordinate of the upper end position and the vertical coordinate of the lower end position of the electric wire 10 '. That is, since the positional relationship between the inspection unit 1 and the camera head 3 is fixed, the distance of the coordinate system in the image is converted into the actual spatial distance by the conversion coefficient set based on this positional relationship, and the line diameter is calculated. Ask. Then, it is determined which type in Table 1 above. In addition, electric wire 1 in the image
The wire diameter of 0'is about 20 to 30 pixels.

(Line Color Judgment) FIG. 7 is a view for explaining the inspection flow for judging the line color. In the line color determination, R, G, B data is sampled for the three electric wires 10 'in the image. However, in addition to the base color and the stripe color, the electric wire 10 may have a silver mark called a band mark as shown in FIG. 7, for example. It becomes impossible to judge. Therefore, as shown by arrows and in FIG.
Data sampling lines are set at intervals wider than the band mark intervals, and sampling is performed at these two points.

First, two pieces of R, G, B data are vertically sampled in the order shown in FIG. 7, and the R, G, B data corresponding to each sampled pixel are converted into color data corresponding to each pixel. To do. This color data is Y indicating the brightness.
Data, S data indicating saturation, and H data indicating hue, which are calculated from R, G, B data based on the following equation (1). It should be noted that R, G, B, Y, and S have a level of 0 to 255, and H has a level of 0 to 360.

[0030]

[Equation 1]

When the R, G, B data of each pixel on the data sampling line is converted into the Y, S, H color data as described above, the color data of only the electric wire 10 'is extracted next. . That is, Y of the data other than the wire 10 'is close to 255 and S is close to 0 in correspondence with the white color. The data is deleted and the color data of only the portion of the electric wire 10 'is extracted. Next, the color data of one data sampling line that can determine the line color, that is, the color data that is not the band mark portion is selected.
Then, as described later, the selected color data is compared with the pixel distribution for each basic color stored in advance to determine the line color.

Since there are a plurality of pixels in the portion of the electric wire 10 'in one data sampling line, the sampled R, G, B data form a data group, and the Y, S, H data converted for each pixel. Also form a group of data. And
When the pixels having the same data value in each data group of Y, S, and H are tabulated, as shown in FIGS. 8 (A), (B), and (C), the normal corresponding to Y, S, and H, respectively. The pixel distribution is distributed. It should be noted that Y, S, and H are dispersed in this way even when the wire color is a single color, even if the lighting is adjusted or the electric wire 10 is used.
Depending on the difference in the reflection angle of the surface of the
This is because the S and H data values are different. Also, the line color is 2
In the case of color, it is naturally dispersed.

Such a set of three pixel distributions can be obtained for each electric wire 10 whose line color is a basic color, and these pixel distributions are at least three of Y, S, and H. When viewed as a set, each basic color is different. If the color area is between the minimum point and the maximum point in each pixel distribution and the color centroid is the most data point, the minimum point data and the maximum point data and the color centroid data representing this color area represent the characteristics of each pixel distribution. There is.

Therefore, the Y, S, and H pixel distributions are obtained in advance for each basic color, and the minimum point data corresponding to each basic color are obtained.
The maximum point data and the color centroid data are registered in the memory in advance. For example, FIG. 9 shows the Y, S, and H pixel distributions for each of the basic colors, red, green, and blue, as an example of the pixel distribution. Actually, there is a pixel distribution for 13 basic colors for each of Y, S, and H, and these data are registered. Then, Y, S, and
The line color is determined by comparing the H data, that is, the unknown data with the registered basic color data.

FIG. 10 is a flowchart showing a line color determination logic to which the color area mapping method in the personal computer 5 is applied. The line color determination logic will be described with reference to FIG. It is assumed that the Y, S, H data of the dummy line to be discriminated has already been obtained as unknown data. First, in step S1, white data is corrected as follows. Although there is white in the line color, white is data other than the line color portion and cannot be detected as a color. Therefore, when the color data of only the portion of the wire 10 'is extracted, the number of pixels corresponding to the wire diameter obtained by the wire diameter inspection is about 3 times (the two wires 10' by the reflection mirror and the wire of the direct image). When the number of unknown data of the line color less than the number of data of 10 ') is obtained, it is determined that the shortage is white and the color data of white is replenished.

By the way, the Y, S, H data of one pixel corresponds to one point in the three-dimensional color space of Y, S, H, and the center of gravity of each of the 13 basic colors (Y, S, H data) is It corresponds to 13 points in the three-dimensional color space. The entire unknown data is distributed around the point corresponding to the color centroid data corresponding to the pixel distribution of the unknown data. Therefore, based on the positional relationship between the unknown data in the three-dimensional color space and the color centroids of the 13 basic colors, the line color is determined by the color centroid distance comparison method and the color area mapping method in step S2 and subsequent steps.

(Color Centroid Distance Comparison Method) First, step S
In step 2, the distances between the points corresponding to the unknown data in this three-dimensional color space and the 13 points corresponding to the respective color centroids of the 13 basic colors (hereinafter referred to as "color centroid distances") are obtained, and The basic color having the minimum color centroid distance is obtained for each unknown data, and the unknown data having the minimum color centroid distance for each basic color (hereinafter referred to as "unknown data belonging to basic color") .) Is counted (the number of pixels).

For example, a set of unknown data (Y for one pixel,
1 set of S, H), if there is unknown data of i-th Y, S, H, D _Yi , D _Si , D _Hi , Y-th, S, H of j-th basic color
_Let G _Yj , G _Sj , G _Hj be the color center of gravity, GL _{j be} the distance of the color center of gravity from the _jth basic color, and C _j be the count value of the number of unknown data belonging to the jth basic color. ) Is calculated.

[0039]

[Equation 2]

Next, the substitution expression of the following expression (3) is executed for j for which GL _j has the minimum value. Note that the formula (3) is C _j
It means that the value obtained by incrementing by 1 is stored in C _j .

[0041]

(Equation 3)

Then, the above calculation and counting are repeated from i = 1 to a.

Next, in step S3, the basic color closest to the entire unknown data in distance is obtained, and the ratio of the unknown data belonging to the basic color to the entire unknown data is obtained. For example, C _OUT1 , C in descending order of C _j
Substituting into _OUT2 and C _OUT3 , the upper 3 colors are determined, and the following equation (4)
Is calculated.

[0044]

[Equation 4]

Next, in step S4, it is determined whether or not the ratio of the closest basic color is 90% or more. If it is 90% or more, the line color of the dummy line is the closest basic color in step S5. It is determined that the color is a single color and the processing is ended. Also,
If it is not 90% or more, the line color is determined by the color area mapping method in step S6 and subsequent steps.

(Color Area Mapping Method) First, in step S6, the upper 3 colors are selected from the ones with the highest proportion, only the unknown data belonging to the basic color corresponding to the upper 3 colors are extracted, and in step S7, The unknown data belonging to the basic colors corresponding to the upper three colors and the color regions of the basic colors of the upper three colors are mapped. For example, the minimum value and the maximum value of the registered color areas of the upper three colors selected by C _OUT1 , C _OUT2 , and C _OUT3 are
_{Y-MIN, k} , C _{Y-MAX, k} , C _{S-MIN, k} , C _{S-MAX, k} , C
_{H-MIN, k} , C _{H-MAX, k} (k = 1,2,3), all unknown data (group b) belonging to the basic color corresponding to the upper 3 colors are D
_Ym , D _Sm , D _Hm (m = 1, 2, ..., B), and the count value of the number of unknown data included in the color region of the kth basic color of the upper 3 basic colors is C _k. , The conditional expression of the following expression (5) is executed.

[0047]

(Equation 5)

Then, the above calculation and counting are performed by k = 1, 2,
3 is repeated from m = 1 to b, and the number of unknown data included in the color region of the basic color is obtained.

Next, in step S8, the order in which the number of unknown data included in the color region of the basic color is large is determined and the ratio thereof is obtained. For example, C _OUT1 in descending order of C _k ,
By substituting the C _{OU T2,} C _OUT3, calculates the following equation (6).

[0050]

(Equation 6)

Next, in step S9, it is determined whether or not the ratio of the basic color having the highest ratio is 90% or more, and the ratio is 90%.
If it is above, it is determined in step S10 that the line color of the dummy line is the basic color having the highest proportion, and the process is terminated. If it is not 90% or more, step S11.
Then, with respect to the line color of the dummy line, it is determined that the basic color with the highest ratio is the base and the basic color with the second highest ratio is the stripe color, and the process ends.

In the above line color determination, the line color is determined by the color area mapping method after the color center of gravity distance comparison method. However, as in the following example, the color area fuzzy mapping is performed after the color center of gravity distance comparison method. The line color may be determined by a method.

FIG. 11 is a flowchart showing the line color determination logic to which the color area fuzzy mapping method is applied. First, in step S1 ', white data is corrected in the same manner as described above, and in step S2', the color centroid distances between the unknown data and the color centroids of the 13 basic colors are obtained in the same manner as described above, and the color of each unknown data is calculated. The basic color having the minimum center-of-gravity distance is obtained, and the number of unknown data (the number of pixels) belonging to the basic color is counted for each basic color. For example, the equations (2) and (3) are calculated in the same manner as described above to count the number (pixel number) C _j of unknown data belonging to the basic color.

Next, in step S3 ', the basic color closest to the entire unknown data in distance is obtained, and the ratio of the unknown data belonging to the basic color to the entire unknown data is obtained. For example, C _OUT1 , C in descending order of C _j
Substituting into _OUT2 , C _OUT3 , and C _OUT4 , the upper 4 colors are determined, and the following equation (7) is calculated.

[0055]

(Equation 7)

Next, in step S4 ', it is judged whether or not the ratio of the closest basic color is 90% or more. If it is 90% or more, the line color of the dummy line is the closest in step S5'. It is determined that the basic color is a single color, and the process ends. If it is not 90% or more, the line color is determined by the color area fuzzy mapping method in step S6 'and subsequent steps.

(Color Area Fuzzy Mapping Method) In this color area fuzzy mapping method, for example, as shown in FIG. 12, the membership function is set in correspondence with the color area of the basic color pixel distribution. This membership function is set so that the grade of the center of gravity of the pixel distribution is "1" and the grades of the minimum and maximum points at both ends of the color area are "0", and the membership function to which unknown data belongs From the unknown data, the grades FUZZY _Y , FUZZY _S , and FUZZY _H belonging to the respective basic color Y, S, and H color regions are obtained. Then, the product of the grades is accumulated for the basic colors, and the line color is determined based on the accumulated value.

First, in step S6 ', all unknown data
C of basic colors that are close to _jWith a high percentage of
Select the top 4 colors from, and select the basic color corresponding to the top 4 colors
Extract only unknown data belonging to, and in step S7 ′,
The unknown data belonging to the basic colors corresponding to the top four colors and
Perform fuzzy mapping with the color areas of the four basic colors
U. For example, C_OUT1, C_OUT2, C_OUT3, C_OUT4Select with
The minimum and maximum values of the registered color areas of the upper 4 colors are
_{Y-MIN, k}, C_{Y-MAX, k}, C_{S-MIN, k}, C_{S-MAX, k}, C
_{H-MIN, k}, C_{H-MAX, k}(K = 1, 2, 3, 4), color centroid
To G_{Y, k}, G_{S, k}, G _{H, k}, Basic colors corresponding to the top 4 colors
All unknown data (group b) belonging to_Ym, D _Sm, D_Hm
(M = 1, 2, ..., B), k of the upper four basic colors
Grade of unknown data included in the color area of the th basic color
The cumulative value of C_{FUZZ, k}And the following equations (8), (9), (1
The conditional expression and calculation expression of 0) are executed.

[0059]

(Equation 8)

[0060]

[Equation 9]

[0061]

(Equation 10)

Then, the above calculation and the accumulation of grades are k
= 1, 2, 3, 4 repeated from m = 1 to b, 4
A cumulative value C _{FUZZ, k} of unknown data grades for the color regions of the one basic color is _obtained .

Next, in step S8 ', the order in which the cumulative value of the unknown data grade is large with respect to the basic color region is determined and the ratio thereof is obtained. For example, the following equation (11) is calculated by substituting C _{FUZZ, k in} the descending order for C _OUT1 , C _OUT2 , C _OUT3 , and C _OUT4 .

[0064]

[Equation 11]

Next, in step S9 ', it is determined whether the ratio of the basic color having the highest ratio is 90% or more, and 90
If it is more than%, it is determined in step S10 'that the line color of the dummy line is the basic color having the highest proportion, and the process is terminated. If it is not 90% or more, step S
At 11 ', the line color of the dummy line is determined to be the basic color having the highest ratio as the base and the basic color having the second highest ratio as the stripe color, and the process is terminated.

As described above, the method for determining the line color is 2
Although the types have been explained, when the determination of the line color is completed, FIG.
The terminal type is determined based on the extracted images shown in (A) and (C). The determination of the terminal type is performed, for example, in FIG.
As shown in (B), a plurality of pixels are vertically scanned from predetermined pixels on the background plate 13 'as shown by arrows to search for the highest luminance data, and a threshold value is obtained from the highest luminance data. The luminance data of each pixel is binarized with this threshold value. Then, based on the binarized data of this pixel,
3 (C) and (D), the periphery of the terminals 20a 'and 20b' is traced, and the position of the change point at this time is determined from the distance information as to the type of the terminals 20a 'and 20b'.

The respective judgment results of the wire diameter, the wire color and the terminal type obtained as described above are output by a CRT or a printer.

In this way, when the line color is determined, the line color is determined by comparing the color data corresponding to each pixel with the pixel distribution of the color data of the basic color, so that the line consisting of the base color and the stripe color is determined. Each color can also be determined by color.

In the dummy line having a stripe,
Since there are two stripes at a position of 180 degrees around, only one side may be imaged by the camera head, but in the above embodiment, the back side of the dummy line A is in two directions by the total reflection plane mirrors 11a and 11b. Since the line color is judged based on the images of the dummy line A for the whole week, the judgment can be made more accurately regardless of the stripe position.

[0070]

As described above, according to the linear material discriminating method of the first aspect of the invention or the linear material discriminating apparatus of the fifth aspect, the surface of the linear material is imaged and the image is taken. R, G, B data of pixels in the image are converted into color data to obtain color data for pixels in the direction orthogonal to the axis of the linear material,
The obtained color data and the pixel distribution in the color data for the basic color are compared to determine the type of the imaged linear material. It is possible to automatically determine the type of linear material due to the difference in color.

Further, according to the linear material discriminating apparatus of the seventh aspect of the present invention, the surface of the linear material is imaged, and the R, G, B data of the pixels in the imaged image is converted into color data. The color data of pixels in the direction orthogonal to the axis of the linear material is obtained, the obtained color data is compared with the pixel distribution in the color data of the basic color, and the direction that intersects the reference line in the image is determined. The pixel is searched for to detect the reference line from the brightness data, and the line diameter of the linear material is determined by obtaining the intersection of the reference line and the linear material from the brightness data along the reference line.
Since the type of linear material is discriminated based on the collation of the color data and the pixel distribution and the determination result of the wire diameter, the difference in the line color and the diameter of the linear material such as an electric wire is targeted. It is possible to automatically discriminate the type of linear material.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a main part of a dummy line determination device to which the present invention has been applied.

FIG. 2 is a perspective view of an inspection unit in the embodiment.

FIG. 3 is a perspective view of an illumination unit according to an embodiment.

FIG. 4 is a diagram showing an example of a captured image of a dummy line in the example.

FIG. 5 is a diagram illustrating an inspection flow for determining a wire diameter in the embodiment.

FIG. 6 is a diagram illustrating a logic for detecting an upper end position of an image of an electric wire according to the embodiment.

FIG. 7 is a diagram illustrating an inspection flow for determining a line color according to the embodiment.

FIG. 8 is a diagram illustrating a pixel distribution of color data in an example.

FIG. 9 is a diagram showing an example of a pixel distribution of color data of basic colors in the embodiment.

FIG. 10 is a flowchart showing a line color determination logic to which the color area mapping method of the personal computer in the embodiment is applied.

FIG. 11 is a flowchart showing a line color determination logic to which the color area fuzzy mapping method of the personal computer in the embodiment is applied.

FIG. 12 is a diagram illustrating a color area fuzzy mapping method according to an embodiment.

FIG. 13 is a diagram illustrating a terminal type determination process according to the embodiment.

FIG. 14 is a diagram showing an example of a dummy line targeted by the determination device of the embodiment of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Inspection part, 2 ... Illumination part, 3 ... Camera head, 4 ... Camera body, 5 ... Personal computer, 11 ... Angled reflection mirror, 11a, 11b ... Total reflection plane mirror, 12
... reference plate, 12a ... reference line.

Claims (9)

[Claims] 1. A linear material having a single color line color among a plurality of predetermined basic colors, or a line color having a single color line color and a plurality of basic colors forming a stripe in the axial direction. A method of discriminating a linear material for discriminating an arbitrary kind of the linear material by determining the linear color of the linear material or the linear material having the line color forming the stripe. That is, the surface of the linear material is imaged, R, G, B data of pixels in the imaged image is converted into color data to obtain color data for pixels in a direction orthogonal to the axis of the linear material. A method of discriminating a linear material, characterized in that the color data is collated with a pixel distribution in the color data for the basic color to determine the line color of the imaged linear material. 2. The linear pattern according to claim 1, wherein the pixel distribution is checked based on the distance between the color centroid of the pixel distribution in the color data of the basic color and the color data of the image. How to determine the material. 3. The frequency of including the color data of the image in the color area of the pixel distribution in the color data of the basic color is calculated, and the comparison with the pixel distribution is performed based on the frequency. The method for discriminating a linear material according to claim 1. 4. A membership function is set in which the center of gravity of the pixel distribution is set to a maximum grade and both ends of the color region are set to a minimum grade in association with the color region of the pixel distribution in the color data of the basic color, and the image is displayed. 2. The method for discriminating a linear member according to claim 1, wherein the color data of the image is compared with the pixel distribution in the color data of the basic color based on the grade of the color data of the membership function. 5. The method of discriminating a linear material according to claim 1, wherein the color data is luminance data, saturation data, and hue data. 6. A linear material having a single color line color among a plurality of predetermined basic colors, or a line color having a single color line color and a plurality of basic colors forming a stripe in the axial direction. A linear material discriminating apparatus which discriminates the kind of a linear material by discriminating the linear color of the linear material or the linear material having the line color forming the stripe. An image pickup means for picking up the surface of the linear material to obtain R, G, B data of pixels in the picked-up image; and a storage means for storing a pixel distribution in color data for the basic color. And converting the R, G, B data into color data to obtain color data for pixels in a direction orthogonal to the axis of the linear material, and calculating the color data and the pixel distribution stored in the storage means. A line color determination unit that determines the line color of the linear material that is collated and imaged And a linear member discriminating apparatus. 7. A linear material having a single color line color among a plurality of predetermined basic colors, or a line color having a single color line color and a plurality of basic colors forming a stripe in the axial direction. A linear material discriminating apparatus which discriminates the kind of a linear material by discriminating the linear color of the linear material or the linear material having the line color forming the stripe. And a reference line arranged on the background of the linear material so as to be orthogonal to the linear material, R of pixels in the captured image by imaging the reference line and the surface of the linear material, An image pickup unit that obtains G and B data, a storage unit that stores a pixel distribution in the color data of the basic color, and both sides of the linear member in the image picked up by the image pickup unit, the reference in the image. Search for a pixel in the direction that intersects the line and From detecting the reference line, find the intersection with the linear member and the reference line searching for pixels along a reference line the detected from the luminance data of the pixel,
Wire diameter determining means for determining the wire diameter of the linear material based on the positions of the intersections on both sides of the linear material; and the R, G, B data converted to color data and orthogonal to the axis of the linear material. Line color determination means for determining color data for pixels in the direction of the line, and comparing the color data with the pixel distribution stored in the storage means to determine the line color of the imaged linear material. An apparatus for discriminating a linear material, which is provided. 8. The camera, wherein the image pickup means is arranged so as to face one side surface of the linear member, and the camera is arranged so as to face a back side surface of the one side surface of the linear member. And a reflecting member for directing a reflected image of the back surface of the linear member into the field of view of the camera by associating a two-sided total reflection mirror with a valley line parallel to the linear member. The linear material discriminating apparatus according to claim 6 or 7. 9. The color data is luminance data, saturation data, and hue data, and the control means performs collation for each pixel distribution in the luminance data, saturation data, and hue data. Item 6 or claim 7
Alternatively, the linear material discriminating apparatus according to claim 8.