JPS62191743A - Method for inspecting leading end part of needle-shaped article - Google Patents

Abstract

PURPOSE:To make it possible to judge the quality of the roundness formed on the leading end part of a needle-shaped article, by calculating the area and circumferential length of the article in the window set to a partial area based on the leading end position of the needle-shaped article to compare the same with standard values. CONSTITUTION:Illumination light is made to irradiate a needle-shaped article 1 being an objective from a lamp 2 and the image in a visual field is picked up by a camera 3 and an image signal is outputted from the camera 3 corresponding to the brightness of each picture element. This image signal is inputted to an A/D converter 4 to be converted to a digital signal, and the image data formed into modulo-(n) system and said image data are binarized on the basis of a threshold value to output a binarized image data. These image data are stored in a modulo-(n) system and binary image memories 7, 8. CPU 10 calculates the leading end of the needle-shaped article from the image data stored in the memory 8 to determine a window and the area and circumferential length of the needle-shaped article in the window are compared with standard values to determine the quality of the roundness of the leading end of the article 1.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for inspecting a needle or a needle-like article having a similar shape, and particularly relates to a method suitable for inspecting the tip of a needle-like article. .

[Conventional technology]

Conventionally, as a method for extracting scratches on special Jζ printed boards, etc., the magazine "Sensor Technology J December 1983 issue (Vo-1,
3, No. 13), pages 91-92, or as described in IEEJ Transactions C50-C8, 49-C8.
This process enlarges the image of the part to be recognized.

Scratches (defects) were detected using a combination of reduction processes.

[Problem that the invention seeks to solve]

The method according to the above-mentioned prior art is effective for inspecting partial scratches such as scratches on printed circuit boards. However, it is not possible to judge the quality of the roundness of a needle-like article, especially whether the tip of the needle-like article has a roundness. Note that the presence or absence of a burr on the needle tip can also be determined using conventional techniques.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for inspecting the tip of a needle-like article, which can determine the acceptability of the roundness of the tip of the needle-like article.

[Means for solving problems]

In order to achieve the above object, the present invention extracts the tip position of a needle-like article from image data, sets a window in a partial area based on the tip position, and calculates the area of the article within the window. By determining the circumference and comparing it with each standard value, it is determined whether the roundness of the tip of the article is within an allowable range, and the quality of the tip of the article is determined.

[Effect]

Since the window is set around the position of the tip of the needle-shaped article, image data of only the portion to be inspected is used. The area of the article obtained from the image data within this window area is quite different depending on whether the tip is fully rounded or if the tip is rounded. The chest circumference of the article obtained from the data similarly varies greatly depending on the shape of the tip.

Therefore, by setting reference values (standard values) and comparing the area and perimeter, the deviation from the standard value can be determined. Acceptance or failure can be determined based on whether or not this deviation is within the allowable range.

[Example] Hereinafter, the present invention will be described in detail based on specific examples.

FIG. 1 shows a system configuration diagram in one embodiment of the present invention, FIG. 2 shows a flow diagram showing its operating method, and FIG.
The reason indicates the image output of the needle tip.

First, in Figure 1, 1 indicates the object to be recognized (needle),
Reference numeral 2 indicates a lamp that illuminates the object 1 with illumination light. A camera 3 captures an image within its field of view including the object. A/
The D conversion unit 4 converts an image signal (analog voltage signal) obtained by imaging an object with the camera 3 into a digital signal of n value (n>3), that is, digital image data, and outputs the converted image signal. . The binarization circuit 6 is the A/D converter 4
The output data is divided by a certain threshold value and binary image data is output. The n-value image memory 7 is A/
The n-value image data from the D converter 4 is stored, and the binary image memory 8 is a memory that stores the binary image data from the binarization circuit 6. 9 is a bus for exchanging data. The central processing unit (hereinafter referred to as CPU) 1o is
Based on the image data processing program stored in the memory 111, the n-value image memory 7 and the binary image memory 81
The stored image data is taken in to calculate the characteristic amounts (ili product, circumference) of the target sea urchin, and it is determined whether the needle is good or bad. njJ processing for recognition (for example, processing such as noise removal and edge enhancement) is also one step in the recognition processing.

The memory 11 stores the above-mentioned image data processing program, and also stores various data necessary for data processing by the CPU 10, and also stores processing results. The input/output control unit (Ilo) 12 performs input/output control. The display unit 13 is composed of, for example, a CRT display device, and displays input data from the keyboard 14, data processing results of the CPU 10 (particularly object recognition processing results, and recognition processing results (good/failure judgment based on the results). result)
, the n value, the contents of the binary image memory 7.8, etc. Keyboard 14 is an operator's tool for inputting data and commands.

The recognition start command generating section 15 is a section that uses a sensor or the like to emit light to indicate the timing of recognition. The needle quality distribution section 16 uses input/output control (Ilo) to determine the quality of the needle tip.
) 12 and is used to sort out non-defective products and defective products. The timing control unit 5 generates various timings necessary for #i positions, such as the timing for storing the image data in the n-value and binary image memory 7.8 in combination with the scanning timing of the camera 3.

Next, the operation of the system shown in FIG. 1 will be explained using FIGS. 2 and 3. First, turn on the system power (
When set to ON), initialization is executed. Various data necessary for the processing are entered manually from the keyboard 14 without referring to the content displayed as guidance on the display section 13. The data is stored in memory 11. Further, at this time, selection between automatic operation and semi-automatic operation is performed by manual key operation. The start timing of the image recognition process is determined by the output from the recognition start command generation unit 15 when the key is pressed as "auto" on the keyboard 14, or by the key-in of the start key on the keyboard 14 when the key is pressed as "semi-automatic". .

The camera 3 captures an image within the field of view of the camera 3 and outputs an image signal according to the brightness of each pixel.

This process is the process of step F1 in FIG.

This image signal (analog voltage signal) is sent to the A/D converter 4
One signal is input, and it is converted into a digital (multi-valued) signal. In this embodiment, n-valued image data and binarized image data obtained by binarizing the image data using a threshold value (this threshold value is set from the keyboard 14 at the time of initialization) are used. Output. This process is the process of step F2 in FIG. The digitized image data is stored in an n-value, binary image memory 7.8. This process is F3 in FIG. Step F1
~The processing of F3 is actually performed continuously at a fixed period,
As a result, the latest image data is stored in the n-value, binary image memory 7.8. CPUl0 is 2
The value is stored in the image memory 8. From the image data, Figure 3 (alfbl(C1((al...normal needle (b))
... Find point A (tip of the needle) of a needle with a rounded tip (C1...needle with one burr on the tip). This process is step F4 in FIG. Next, from the tip of the needle (point A),
CPUl0 or processing area (hereinafter referred to as window)
Determine. The actual window generation position is shown in Figure 3 fa.
l(bl(shown in cl). The window size is set using the keyboard 14 at the time of initialization. This process is step F5 in FIG. 2.
PU10 is the area 81 of the button in the window and the perimeter L1
seek. This process is performed at step F6 in FIG. It is F7. The CPU10 has an area value S1 and a standard area value So (this value may be set at initialization to a value determined to be appropriate through past recognition experiments, or when object recognition is performed continuously. It is sufficient to memorize the optimal value obtained through previous learning and use it this time.

) and find the difference ΔSa. This process is step F8 in FIG.

Then, this △Sa is within a preset tolerance range ±εN
Determine whether it is within the range. This process is step F9 in FIG. In this judgment, if the needle deviates from △Sa or ±εAl, the needle is determined to be defective, and the input/output control unit (Ilo) 12 sends the needle pass/fail to the unit 16 to remove it from the line. Outputs the directives to be excluded. on the other hand,
In the determination at step F9, if ΔSa is within ±ξA1, the process proceeds to step Fil, where the peripheral length L1 and the standard value are determined.

(calculated in the same way as the standard area value So) △L
seek. This △L is the preset tolerance range ± and A
Determine whether it is within 1. This process is step F12 in FIG. In this judgment, if ΔL deviates from ±εA1, the needle is determined to be defective in the same way as if it deviates from the area tolerance range ±εA1.
The input/output control unit (Ilo) 12 issues a command to exclude it from the line. Regarding the roundness of the tip of the sword, even if it is just the area 2 the perimeter, the perforation position of the window as shown in Figure 3 (al (bl) or △ × is different, so the difference from the standard value (△S
a, ΔL) is thick (can be determined).

On the other hand, the burr on the needle tip is the difference from the standard value (△Sa).

△L) is small and difficult to understand. Then, the burr on the needle tip is in step F9. If it is not found in step F12, reduction processing is performed. For example, if an image like (ci) in Figure 3 is reduced only within the window, it will become like (bl), the entire image will be reduced and the burrs will disappear.This process is performed in step F13 in Figure 2. Next, when you enlarge it, it will look like (C), and the burr part will disappear.
It becomes a normal sword tip. This process is slap F14 in FIG. The surface fjts2 of the needle tip in this state is determined. This process is step F15 in FIG. Face M
The difference Δsb between S2 and the area S1 determined in step F6 is determined. This process is step F16 in FIG. Third
If there is a burr as shown in Figure (C1), it can be determined as the area of the burr part in △Sb.In addition, in the case of a normal needle as in Figure 3 (a), since it is o in △sb, It is possible to use a very small value for the preset tolerance range ±εM. ((The area of the burr part △sb is
Since this is a small value, if the window area is large (the recognition area is large), the effects of noise and unstable pixels will be reduced.
It is possible that this could become a problem.

In this case, when checking for burrs, you can set the window small so that you only see the burrs. That is, in step F5, another window for checking blockage is generated, and in step F7, its area is also calculated, and in that window, reduction → enlargement (Stella 713, 14) is performed, and △sb is obtained. good. )) If it is △sb or outside the tolerance range ±εAfe, it is determined that it is a defective product, and the input/output control unit (Ilo) outputs a command to exclude it from the line for the needle pass/fail distribution. . On the other hand, if △sb is within ±ε,
It is judged to be a good product and the lines are applied as is.

According to this embodiment 1, it is possible to allow the needles flowing through the line to pass through the needles as they are if they are good, and to exclude defective products.

〔Effect of the invention〕

According to the present invention, the rounded state of the needle tip can be recognized from the needle tip area and peripheral length within the window, so it is suitable for inspecting the needle tip.

[Brief explanation of drawings]

Fig. 1 is a system configuration diagram showing an embodiment of the present invention;
The figure is an operation flow diagram of Embodiment I in FIG. 1, and the third factor is an image of each needle tip. 1...Object (needle), 2...Lighting, 3
...Camera, 4...A/Di converter, 5
...timing control unit, 6...binarization circuit, 7...n-value image memory, 8...
Binary image memory, 9... Bus, 10...
・Central processing unit, 11...Memory, 12...
...Input/output control section, 13...Display section, 14...
...Keyboard, 15...Recognition start command generation unit, 16...Second factor diagram for needle quality distribution (C) (d) (e)

Claims (1)

[Claims] 1. Take an image of the needle-shaped article to be inspected, convert it to digital image data, set a window in a partial area of the image data based on the tip position of the article, and calculate the area and size of the article within the window. 1. A method for inspecting the tip of a needle-shaped article, characterized in that the circumference of the article is determined and the roundness of the tip of the article is judged by determining the perimeter and comparing it with each standard value.