JPH05296724A - Parts recognition method - Google Patents

Abstract

PURPOSE:To provide a method for detecting the position of an electrode as a recognition object at high speed and with accuracy, regarding image recognition and processing in an electronic part packaging facility or the like. CONSTITUTION:With both lights for transmission and reflection turned on, an electronic part is photographed and an image pattern is obtained at the first process. Then, a boundary between an electronic part pattern in the image pattern and a background is obtained, and the outline of the part is thereby recognized. The rough inclination and center of the part is obtained on the basis of the outline so obtained. A processing area for detecting an electrode is set, according to the rough center and inclination, and the position of the electrode is detected by finding the extreme value of brightness projection data in the processing area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image recognition method of a visual recognition device required for mounting electronic parts at high speed and with high accuracy in electronic part mounting equipment and the like. The present invention relates to a method of accurately measuring the electrode position of the electrode according to an image pattern.

[0002]

2. Description of the Related Art In recent years, in the field of mounting electronic components, a technique for mounting electronic components on a circuit board at high speed and with high precision is required. Therefore, there is a tendency to utilize an image recognition technique that processes an image pattern obtained by capturing an image of an electronic component at high speed to detect the position and inclination of the electronic component at high speed and accurately. In addition, as electronic components are becoming finer, the number of pins of electrodes is increasing. Therefore, for electronic components having a large number of electrodes, it is desired to detect the position of each electrode.

Conventionally, as a method of recognizing an electronic component utilizing image recognition technology, a transparent recognition method of illuminating the target electronic component from behind and processing its silhouette to detect the center, inclination, electrode position, etc., A reflection recognition method has been proposed in which light is applied from the front and the light reflected from the electrodes is processed to detect the center, inclination, electrode position, and the like.

A conventional component recognition method will be described below with reference to the drawings.

FIG. 4 is an example of a transmission recognition method and shows an example of detecting the tip position of the lead by illuminating the leaded electronic component 42 from behind (hereinafter referred to as transmission illumination). As shown in FIG. 4A, when the transillumination 41 is applied to the leaded electronic component 42 and an image is taken by the video camera 43, an image pattern 44 as shown in FIG. 4B is obtained. The silhouette of the image pattern 45 of the leaded electronic component 42 is displayed by transmitted illumination. The boundary between the image pattern 45 of the leaded electronic component 42 and the background is shown in FIG.
When the locus of the boundary tracking is obtained by the detection as shown in (c) of FIG. 3, the position 46 at which the U-turn is made is detected. The position 46 at which this U-turn is made is detected as the electrode position to be detected. Since this transparent recognition method processes silhouettes,
The logic is simple and the processing is fast.

However, with the transmission recognition method, it is difficult to detect the electrode position of an electronic component (referred to as a J-lead electronic component) in which the electrode is bent inward from the outer shape of the component. Therefore, a reflection recognition method has been proposed that can directly capture the electrodes.

FIG. 5 shows an example of the reflection recognition method, in which the lead position of the J-lead electronic component 51 is detected. As shown in FIG. 5A, the J lead electronic component 5
Illuminate (hereinafter referred to as reflective illumination) 52 from the front of 1
When picked up by the video camera 43, an image pattern 53 as shown in FIG. In the image pattern 54 of the J-lead electronic component 51, the reflected light of the reflective illumination 52 is projected, so that the electrode portion is bright. FIG. 5 (c)
As shown in FIG. 6, the electrode portion 55 that shines brightly is scanned in the manner of S, and the rough position is detected. The processing window 56 is provided on the basis of the detected rough position, the luminance distribution in the processing window 56 is projected as 57, and the extreme value 58 is detected. Since the electrode portion has high brightness, the detected extreme value 58 is detected as the electrode position to be detected.

[0008]

As described above, as the number of pins of electronic parts increases, the limitation of arranging electrodes on the side surface of parts is being reached. Against this background, Figure 6
As shown in Fig. 3, some parts have been introduced in which the electrodes are arranged inside the outer shape of the part.

As described above, in the transmission recognition method, it is impossible to recognize the electrodes arranged inside the outer shape.

Also in the reflection recognition method, when the component 61 is inclined as shown in FIG. 6A, since the electrodes are inside the outer shape of the component, if the scanning S is performed to detect only the electrodes, The wrong processing window 62 is set. Further, if the outer shape is scanned S, in the case of the component 63 having the shape as shown in FIG. 6B, a wrong processing window 64 as shown may be set. In either case, it is possible to solve the problem to some extent by making the scanning interval dense and widening the range, but it cannot be said to be a good solution because the processing speed becomes considerably slow.

An object of the present invention is to overcome the above-mentioned drawbacks of the prior art and to provide a high-speed and highly reliable component recognition method.

[0012]

In order to solve the above problems, the present invention uses the following means.

First, an image pattern is obtained by picking up an image of a target electronic component while simultaneously applying transmitted illumination and reflected illumination (first step). Since the background has a light intensity due to transmitted illumination,
The boundary between the component body and the background is detected (second step). From the obtained boundary, the approximate center and inclination of the component are detected (third step). Next, a processing area for electrode detection is set using the detected center and inclination, the luminance distribution in the processing area is analyzed, and the position of the electrode is detected (fourth step).

[0014]

The present invention can make use of the advantages of both the transmission recognition method and the reflection recognition method by the above-mentioned configuration.
That is, it is possible to easily and accurately set the processing area of the reflection recognition method that enables highly accurate recognition by using the high-speed tilt / center detection in the transmission recognition method.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS As an embodiment of the present invention, a method for detecting the electrode position of a PGA (Pad Grid Array) component in which electrodes are arranged inside the outer shape of the component will be described with reference to FIGS. 1, 2 and 3. ..

In FIG. 1, the first step is an image pattern input step S1. As shown in FIG. 2A, the PGA component 21, which is the recognition target, has a transmission illumination 22 and a reflection illumination 2.
3 is applied, and an image is obtained by an image pickup means such as the video camera 24 to obtain a video signal.

The obtained video signal is digitized to form an image pattern 25 shown in FIG. 2 (b).

The second step is a component outline detection step S2. In the image pattern 25, the background portion has a high brightness due to transmitted illumination. Therefore, a significant change in the brightness level is seen at the boundary between the component body and the background. Therefore, for the image pattern 25 obtained in the first step, as shown in FIG.
As shown in (a) of (1), the tracking P of the boundary point where the luminance change is remarkable is performed. The boundary thus obtained is detected as the outer shape of the PGA component 21.

The third step is center / tilt coarse detection S3. The center / inclination 32 is determined from the external shape of the component detected in the second step by using a conventionally used method. Various methods have been proposed as a method of calculating the center / inclination from the outer shape, but for example, a method of obtaining a circumscribed rectangle of the outer shape of the component and setting the center as the center of the component,
There is a method of detecting the inclination by calculating the second moment between a certain arbitrary point and the boundary point group forming the outer shape.

The fourth step is an electrode position detecting step S4. A processing window 33 as shown in FIG. 3B is set by utilizing the rough center and inclination determined in the third step, and the external dimensions and inter-electrode dimensions given in advance. The brightness distributions in the set processing window 33 are projected in the α and β directions, and the projection data 34, 34
To get The intersection 35 of the extreme positions of the obtained projection data 34, 34 is detected as the position of each electrode.

The process in the fourth step is repeated to detect all the electrode positions, and the process is completed.

[0022]

According to the present invention, since the processing area can be set by utilizing the center and inclination coarse detection, which makes use of the advantages of the transmission recognition method, the disadvantages of the reflection recognition method can be corrected. The processing area can be set easily and accurately. According to the present invention, by utilizing the advantages of each of the transmission recognition method and the reflection recognition method, the electrode position of a more reliable target can be obtained without being affected by the shape and orientation of the electronic component to be processed. Can be detected.

[Brief description of drawings]

FIG. 1 is a flowchart showing an outline of an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an imaging method and an image pattern according to an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a method for detecting an electrode position according to an embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a method of detecting electrode positions of an electronic component of a conventional example.

FIG. 5 is an explanatory diagram showing another conventional electrode position detection method.

FIG. 6 is an explanatory diagram showing a problem of the conventional example.

[Explanation of symbols]

 21 PGA component (recognition target) 22 Transmitted illumination 23 Reflective illumination 24 Video camera 32 Center / tilt of component 33 Processing window (processing area) 35 Electrode position

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification number Reference number within the agency FI technical display location H05K 13/08 B 8315-4E (72) Inventor Masao Iriya 1006 Kadoma, Kadoma, Osaka Prefecture Matsushita Electric Sangyo Co., Ltd.

Claims (1)

[Claims] 1. A first step of capturing an image pattern of a recognition target object while illuminating the recognition target object from both front and rear surfaces, and detecting the outer shape of the recognition target object from the boundary between the image pattern of the recognition target object and the background. A second step; a third step of detecting a rough center and an inclination of the recognition object from the outer shape of the recognition object detected in the second step; and a rough recognition object of the recognition object detected in the third step. A component recognition method comprising a fourth step of setting a processing area using a center and an inclination, and detecting an electrode position of a recognition target object based on a luminance distribution in the processing area.