JPH0672780B2 - Defective product judgment image processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION A. Field of Industrial Application The present invention relates to a defective product determination image processing apparatus, and is suitable for use in inspecting the quality of a product having a complicated shape, in particular, the defect defect of parts.

B. Summary of the Invention The defective product determination image processing apparatus of the first invention captures a plurality of non-defective work pieces with a TV camera in advance, binarizes the image signal of the TV camera, counts the number of pixels of the non-defective work pieces, and counts the count value. Is statistically processed to find the average value μ and standard deviation σ, and μ
± 3σ is stored as a determination value, the inspection work is then photographed by the same TV camera, the image signal is binarized to count the number of pixels of the inspection work, and the number of pixels of the inspection work deviates from the determination value. Then, the inspection work is determined to be defective.

The defective product determination image processing device of the second invention photographs a plurality of non-defective works in advance with a TV camera, binarizes the image signal of the TV camera to count the number of pixels of non-defective works, and statistically processes the count value. Obtain the average value, store the average value as a reference value, then take an image of the inspection work with the same TV camera, binarize the image signal, and count the number of pixels of the inspection work. If there is a difference in the number of pixels, the inspection work is determined to be defective.

C. Conventional technology One of the representative examples of products with complicated shapes is a wiring harness for automobiles.

A wide variety of electrical components are used in automobiles, and there are many electrical components with relatively low current capacity that perform lighting, signals, control, warnings, charging, measuring instruments, window wiping, and the like.

A wiring harness is a wiring harness in which these electrical components are connected in advance so that they can be smoothly assembled on an automobile assembly line. The wiring harness is constructed by bundling a large number of electric wires cut to a required length with vinyl tape or the like, and attaching harness parts (tube, clamper, grommet, protector, taping, etc.) to it.

In the usual manufacturing method, the wiring harness sets the focusing wire on the drawing board (the one with the drawing for production attached), and while sending the drawing board along the manufacturing line, It is made by installing the necessary harness parts.

The wiring harness produced in this way is inspected whether all necessary parts are attached, and if any parts are missing, it is determined as a defective product.

Conventionally, as a method for inspecting missing parts, a visual inspection by human eyes has been carried out, and the introduction of visual inspection by image processing has been considered.

A conventional visual inspection by image processing will be described with reference to FIG.

As shown in FIG. 8, the wiring harness 1 is set on the drawing board 2, and the drawing board 2 is fed in the direction of arrow A. 1a is a focusing electric wire, and 1b is a harness component. And
The industrial television camera 3 photographs the wiring harness 1 and the drawing board 2, and sends an image signal to the image processing device 4.
The image processing device 4 has a built-in television camera interface and a microcomputer, and each photographing region obtained by dividing the surface of the drawing board 2 into a plurality of portions corresponding to the photographing field of view of the television camera 3, for example, a dashed line 4 in the figure. Imaging areas I, II, I
The image processing device 4 captures the image signal of the television camera 3 at each timing when the photographing field of view of the television camera 3 is matched with II and IV.

In the image processing device 4, the terminal 5 (having a keyboard and a CRT) is used to register an image showing a position where each harness component is normally occupied and a regular shape of each harness component in association with each photographing area. ing. A monitor 6 is connected to the image processing device 4.

Then, the image processing device 4 determines each harness component 1b of the wiring harness 1 to be inspected based on the captured image signal.
The position and shape of is detected, and the detected actual position and shape,
By comparing the registered image with each harness part
Determine if 1b is installed correctly.

D. Problems to be Solved by the Invention By the way, of the methods for inspecting missing parts of wiring harnesses, the visual appearance inspection is inferior in inspection accuracy, and the operator is very tired.

On the other hand, in the conventional visual inspection by image processing, since the regular positions and shapes of all the harness components 1b must be registered as images, it is necessary to register a huge amount of data, and it takes time to register. It takes. Further, it takes time to inspect the position and shape of the harness component 1b of the wiring harness 1 to be inspected based on the image signal, and it is difficult to correctly detect the position and shape. For these reasons, the inspection cannot be performed easily and quickly.

The present invention has been made in view of the above-described conventional art, and an object of the present invention is to provide a defective product determination image processing apparatus that can perform a missing inspection automatically and accurately without using visual inspection.

E. Means for Solving the Problem The defective product determination image processing apparatus according to the first aspect of the present invention outputs a TV camera which is commonly used in the same field of view for shooting a non-defective work and an inspection work and a teaching command and an inspection command by switching. The image processing apparatus includes an apparatus and an image processing apparatus that inputs an image signal from a television camera to determine whether or not an inspection work is defective, and the image processing apparatus is configured to detect an image signal of the work from the television camera as a shadow image. A binarization processing unit that divides into pixels and work pixels, a counting unit that counts the number of work pixels, and when the teaching command is output, input the count value of non-defective work pixels and input the average value μ Standard deviation σ
And a statistical processing unit that outputs μ ± 3σ as a determination value, a memory that stores the determination value, and a determination value that stores the count value of the pixel of the inspection work in the memory when the inspection command is output. It is characterized in that it has a judging section for comparing and judging that the inspection work is defective when it is out of the judgment value.

The defective product determination image processing apparatus according to the second aspect of the present invention includes a television camera that is commonly used for capturing non-defective workpieces and inspection workpieces in the same field of view, an apparatus that switches between a teaching instruction and an inspection instruction, and outputs an image from the television camera. And an image processing device for determining whether or not the inspection work is defective by inputting a signal, wherein the image processing device converts the image signal of the work from the television camera into a pixel of the back shadow and a pixel of the work. A binarization processing unit that separates, a counting unit that counts the number of pixels of the work, a statistical processing unit that inputs the count value of the pixels of the non-defective work when the teaching command is output, and an average processing unit, and an average The memory that stores the value as the reference value is compared with the reference value stored in the memory when the inspection command is output and the pixel count value of the inspection work, and the difference is set. It has a judgment part which judges that the inspection work is defective when it is larger than the value.

F. Action Since the area of a non-defective work is different from that of a defective work with missing parts, when the same TV camera captures the same field of view, there is a difference in the number of pixels of the work obtained by binarizing the image signal that exceeds the error range. Occurs.

Therefore, in the apparatus of the first invention, first, as teaching, a plurality of non-defective works are photographed in advance by a TV camera, the image signals thereof are binarized to count the number of pixels of the non-defective works, and the count values are statistically processed and averaged. The value μ and the standard deviation σ are obtained, and μ ± 3σ is stored as a judgment value. Next, at the time of inspection, the inspection work is photographed by the same TV camera, the image signal is binarized, and the number of pixels of the inspection work is counted. The number of pixels of the inspection work is compared with the determination value. If the number of pixels deviates from the determination value, it is determined that the inspection work is defective. In this way, the count value of only good products is statistically processed and the judgment value μ
By setting ± 3σ. This is advantageous in that it is not necessary to prepare a large number of defective samples, as compared with the case where a good sample and a defective sample are measured and an intermediate value between them is used as a determination value. For example, when there are many inspection points such as a visual inspection of the wire harness, it is troublesome and wasteful to intentionally cause a defective state at each inspection point. Also,
Since defective samples are measured, the number of samples is large and the time required to determine the determination value is long.

In the device of the second invention, first, as teaching,
A plurality of non-defective works are photographed with a TV camera, the image signals are binarized, the number of pixels of non-defective works is counted, the count value is statistically processed to obtain an average value, and the average value is stored as a reference value. . Next, at the time of inspection, the inspection work is photographed by the same TV camera, the image signal is binarized, and the number of inspection work pixels is counted. The reference value and the number of pixels of the inspection work are compared, and if there is a difference, it is determined that the inspection work is defective. In this case, since the reference value is determined by statistical processing, variations in non-defective work hardly affect the determination result.
Moreover, since the allowable range of the difference is determined by the set value, the determination accuracy can be freely determined by the set value.

G. Examples Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a system system of an embodiment of the present invention, FIG. 2 is a plan view showing a transportation line of a board, FIG. 3 is a block diagram of a board, and FIG. 4 is a timing chart of operation. FIG. 5 is a flowchart of teaching, FIG. 6 is a flowchart of statistical processing, and FIG. 7 is a flowchart of inspection.

As shown in FIG. 3, the drawing board 10 is obliquely installed on the carriage 11, and the carriage 11 runs along a guide rail 12 by a power source (not shown). The drawing board 10 is white. 10a is a harness support member, and 11a is a wheel. Therefore, as shown in FIG. 2, the drawing board 10 is sent in the direction of the arrow B, and cyclically moves at the line speed V between the assembly line 13 and the inspection line 15.
Then, while the drawing board 10 moves along the assembly line 13, the wiring harness 14 (see FIG. 2) is assembled on the drawing board 10. Reference numeral 14a is, for example, a black focused electric wire, and 14b is a harness component of a color other than white such as blue and red. The assembled wiring harness 14 is inspected for missing parts while moving along the inspection line 15. A specific explanation of the inspection will be given later. The wiring harness that has passed the inspection is removed from the drawing board 10, and the drawing board 10 returns to the assembly line 13 again.
The wiring harness determined to be defective is returned to the assembly line 13 while being attached to the drawing board 10 and is reworked.

Further, as a pretreatment (teaching) for the inspection on the drawing board 10, it has already been confirmed by visual inspection that the non-defective work, that is, all the necessary harness components 14b are correctly attached to the focusing electric wire 14a. Wiring harness G
Is installed. A plurality of different good works G are used because of the statistical processing.

In this embodiment, four industrial TV cameras TV1, TV2, TV3, TV4 are used in order to increase the resolution according to the size of the work. As shown in FIGS. 1 to 3, in the inspection line 15,
A support member 16 is installed diagonally above the drawing board 10, and each of the TV cameras TV1, TV2, TV3, TV4 is installed on the supporting member 16 and
Faced to 0. The field of view of each TV camera is adjusted according to the inspection accuracy, but since it was set to be much narrower than the drawing board 10, the drawing board 10 is vertically covered by shifting it in the direction C perpendicular to the board traveling direction B. I am trying to do it.
Further, to cover the drawing board 10 in the horizontal direction, marks 17 are provided in the drawing board 10 at intervals in the horizontal direction, and the image processing apparatus is provided each time the mark 17 is detected by a photoelectric sensor or the like provided in the inspection line 15. 20 is designed to capture the image signal of the TV camera. Further, the four TV cameras TV1 to TV4 are shifted in the direction B of the drawing board so that image processing is sequentially performed for each TV camera, and a sensor is provided corresponding to each TV camera to provide four sensors PH1, PH2, PH3 and PH4 are also staggered. Further, a limit switch LS is provided in the inspection line 15 to detect that the drawing board 10 has reached the inspection line 15 and give a start command to the image processing apparatus 20.

As a result of the above, the drawing board 10 is divided into a plurality of photographing areas such as ..., As shown in FIG.
Each time PH1 to PH4 detect mark 17,
The image processing of each photographing area is performed in the order of.

Reference numeral 18 is a light source for illumination such as a fluorescent lamp, and BS is an emergency stop switch. In addition, the inspection line 15 has a dark curtain to block external light.
Covered with 19.

As shown in FIG. 1, the image processing device 20 is supplied with analog image signals S1, S2, S3, S4 from each of the television cameras TV1 to TV4, and the mark detection signals M1, M2, from each of the sensors PH1 to PH4. M
3, M4, start command signal 21 from limit switch LS
However, the emergency stop signal 22 is sent from the switch BS and the teaching command and the inspection command are sent from the switch SW by switching the command signal 23.

The image processing device 20 includes a binarization processing unit, a counting unit, a statistical processing unit, a memory, and a determination unit, and a switch.
By operating the SW, the teaching based on the non-defective wiring harness G shown in FIGS. 5 and 6 and the inspection of the wiring harness 14 to be inspected shown in FIG. 7 are performed.

Each functional unit of the image processing device 20 is realized by a microcomputer and appropriate hardware and software.

The binarization processing unit divides the input image signals S1 to S4 into black pixel signals indicating the wiring harnesses G and 14 and white pixel signals indicating the drawing board 10 and binarizes each time the marks are detected by the sensors PH1 to PH4. And send it to the counting section. In order to reduce the processing time, when the image signal of a certain television camera is captured, the preparation for capturing the image signal of the next television camera is immediately performed. Also,
In order to adjust the area to be binarized, the targets 24 are provided on the plate 10 at intervals in the lateral direction, and correction is performed so as to binarize only the image signal of a predetermined area based on the target 24. Is going.

The counting unit counts the number of pixels of the black pixel signal in the binarized image signal for each photographing region, ... And sends the count value to the statistical processing unit, the memory, and the determination unit.

When the teaching command is given from the switch SW, the memory temporarily saves the count value from the count section to the address corresponding to the shooting area for statistical processing, and also stores the calculation result in the statistical processing section in the affected area. Save to the corresponding address.

As an embodiment of the first aspect of the invention, when the statistical processing section receives a teaching command from the switch SW, it inputs the count value from the memory for each photographing area, and calculates the average value μ and the standard deviation (shown in FIG. 6). The judgment value μ ± 3σ is calculated from the calculation of (variance) σ and stored in the memory. However, n is the number of times of sampling, xn is the count value of the nth sample, μn is the previous average value, μn _{+ 1} is the current average value, and the count value is the average value as it is at the first time. Further, ▲ σ ² _n ▼ is the previous dispersion, and ▲ σ ² _{n + 1} ▼ is the current dispersion.

In the second invention, the statistical processing unit only calculates the average value μ in FIG. 6 and stores it in the memory with μ as a reference value.

As an embodiment of the first aspect of the present invention, the determination unit compares the count value from the count unit with the determination value μ ± 3σ stored in the memory for each imaging region when the inspection command is given from the switch SW, and As a result, when it is out of the determination value in any of the areas, it is determined that the harness component is missing or excessive, and the determination signal 25 indicating "defective" is output.

In the second invention, the determination unit compares the count value from the count unit and the reference value μ stored in the memory for each imaging region when the inspection command is given from the switch SW, and the comparison result indicates which one of them. If the difference between the two is greater than the set value (this is determined in advance according to variations in the size of harness parts, judgment accuracy, etc.), it is determined that the harness parts are missing or extra. Then, the determination signal 25 indicating "defective" is output.

From the above, in teaching, a plurality of non-defective samples are used, non-defective wiring harness G is set on each drawing board 10 and sequentially sent to the inspection line 15.
Each time PH1 to PH4 marks are detected, the image signals of the TV cameras TV1 to TV4 are captured, the number of black pixels is counted, and the count value is taken as the area in each shooting area of the wiring harness G, and the judgment value μ ± 3σ Alternatively, teach the reference value μ.

Next, in the inspection, the wiring harness 14 assembled on the drawing board 10 is sent to the inspection line 15, and the image signal is detected each time the marks of the sensors PH1 to PH4 are detected without changing the visual fields of the TV cameras TV1 to TV4. The number of black pixels is captured, counted, and compared with the judgment value μ ± 3σ or the reference value μ of the number of black pixels taught for each photographing area to judge whether or not there is a defect. If it is defective, rework is performed.

A personal computer 26 having a switch SW on a keyboard is connected to the image processing apparatus 20 and is used for operating the image processing numerical value 20. When it is determined that the component is missing, the result of this determination is displayed on the CRT display device 27. Further, a monitor 28 is connected to the image processing device 20 for adjustment of image processing so that images taken by each of the television cameras TV1 to TV4 are displayed on the monitor 28, and further recorded on a recording device (VTR) 29. A record is kept so that the examination contents can be investigated at a later date. In FIG. 1, reference numeral 30 is a power supply unit, which contains a power supply for a camera and a high frequency power supply for a fluorescent lamp.

After all, in the present embodiment, a plurality of non-defective wiring / harnesses G having no missing parts are used, the judgment value or the reference value is obtained by the statistical processing from the number of black pixels of each non-defective product, and these values and the wiring to be inspected. The presence / absence of a component is determined by comparing the number of black pixels indicating the harness 14 by simple area comparison by binarization. Therefore, the determination accuracy is high. Further, the processing of the image processing device 20 is only a simple process such as binarization, counting, statistical processing, and comparison, so the processing speed is extremely fast, and therefore the moving speed of the drawing board 10 can be increased, and the inspection efficiency is improved. Better and can perform inspections at online speeds.

The present invention can be applied not only to wiring harnesses but also to inspect for missing parts in other inspection works.
In this case, no matter how many shapes the inspection work has, or even if it is difficult to detect individual parts, there is no problem because the missing parts are determined by area comparison. Further, the field of view of the television camera may be adjusted according to the inspection accuracy, and therefore one to many television cameras may be used according to the inspection accuracy and the size of the work.

H. Effect of the Invention According to the present invention as specifically described in connection with the embodiments above, the number of pixels indicating a non-defective work is binarized by binarizing the image signal obtained by photographing the non-defective work and the inspection work with the same TV camera. From these, a determination value or a reference value is obtained by statistical processing, and these values are compared with the number of pixels indicating the inspection work to determine whether or not there is a missing part. Since the judgment value or the reference value is determined by the statistical processing in this way, the error is small. Further, since the determination is made by the simple area comparison, it is not necessary to teach the component position, dimension, etc. of the work component, and the inspection can be performed by a quick and simple process. Also, when the type of work to be inspected is changed, it is only necessary to regard that type of work that has been confirmed to be non-defective as a non-defective work.
No change to the inspection program is required.

[Brief description of drawings]

FIG. 1 is a system diagram showing a system system according to an embodiment of the present invention, FIG. 2 is a plan view showing a transportation line of a drawing board, FIG. 3 is a configuration diagram showing a transportation state of a drawing board, and FIG. Is a diagram showing operation timing, FIG. 5 is a flowchart of teaching,
FIG. 6 is a flow chart of statistical processing, FIG. 7 is a flow chart of inspection, and FIG. 8 is a system system diagram showing a conventional technique. In the drawings, 10 is a drawing board, 14 is a wiring harness as an inspection work, G is a wiring harness as a non-defective work, SW is a switch for teaching and inspection command change, TV1, TV2, TV3, TV4 are TV cameras, Reference numeral 20 is an image processing device.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tsuneetsu 1-10-28 Nagano, Inazawa-shi, Aichi Prefecture (72) Inventor Kiyohide Abe 5 Oshibana, Nishibiwajima-cho, Nishikasugai-gun, Aichi Prefecture (56) References JP-A-61 -169749 (JP, A) JP 62-85803 (JP, A) JP 61-194305 (JP, A) JP 62-180251 (JP, A)

Claims (2)

[Claims] 1. A television camera which is commonly used for photographing non-defective workpieces and inspection workpieces in the same field of view, a device for switching and outputting a teaching command and an inspection instruction, and an inspection workpiece which receives an image signal from the television camera. An image processing device for determining whether or not there is a defect, the image processing device, a binarization processing unit that divides the image signal of the work from the television camera into pixels of the back shadow and pixels of the work, The counting unit that counts the number of pixels of the work and the count value of the pixels of the non-defective work when the teaching command is output, and the average value μ and standard deviation σ
And a statistical processing unit that outputs μ ± 3σ as a determination value, a memory that stores the determination value, and a determination value that stores the count value of the pixel of the inspection work in the memory when the inspection command is output. A defective product determination image processing apparatus, comprising: a determination unit that determines that the inspection work is defective when compared and deviates from the determination value. 2. A television camera which is commonly used for photographing non-defective workpieces and inspection workpieces in the same field of view, a device which switches between a teaching instruction and an inspection instruction and outputs, and an inspection workpiece which receives an image signal from the television camera. An image processing device for determining whether or not there is a defect, the image processing device, a binarization processing unit that divides the image signal of the work from the television camera into pixels of the back shadow and pixels of the work, A counting unit that counts the number of pixels, a statistical processing unit that calculates the average value by inputting the count value of the non-defective workpiece pixels when a teaching command is output, and a memory that stores the average value as a reference value. When the inspection command is output, the reference value stored in the memory is compared with the pixel count value of the inspection work. If the difference is larger than the set value, the inspection work is defective. A defective product determination image processing device, comprising: a determination unit that determines that there is a defective product.