JP2516844B2 - Parts detection method and device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a component detecting method and device for detecting the position of a component by image processing.

[0002]

2. Description of the Related Art Conventionally, as one of methods for detecting the position of a minute component such as an electronic component by image recognition, an image processing area is divided into a plurality of small areas in advance, and an object is first detected by rough scanning. There has been proposed a method of recognizing an area to be processed as an area exceeding a certain image density value, and then newly densely scanning only the small area to quickly detect the position of the component (Japanese Patent Laid-Open No. HEI-1). -240987 gazette).

[0003]

However, in the above-mentioned conventional method, since the size (area) of the small area to be densely scanned is determined in advance, it is actually necessary to scan even in the small area. There is a problem in that even a part without a part (a part where no part exists) is densely scanned, resulting in wasted processing time.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a component detecting method and device capable of quickly detecting the position of a component by shortening the image processing time. Especially.

[0005]

According to the method of the present invention, an area including at least one end of a part is designated on a processing screen,
The position of one end of the component is detected by scanning the image in the region, and the region including the other end of the component is determined based on the position information of the one end of the component and the outline information of the component, The image in that area is scanned to detect the position of the other end of the component, and the next end of the component is included based on the position information of the component end detected so far and the outline information of the component. It is characterized in that the area is determined, and thereafter all the end positions of the parts are detected in the same manner.

Further, the apparatus of the present invention comprises an image pickup means for picking up an image of a part, an end detecting means for detecting an end of the part by scanning an image obtained by the image pickup means in a designated area, and Processing area determining means for determining the area on the screen to be detected next by the edge detecting means based on the position information of the edge of the component detected by the edge detecting means and the outline information of the component; It is characterized by including a positioning output means for calculating the position of the component by scanning the image based on the position information of the component end obtained by the means, and outputting the result.

[0007]

According to the present invention, the processing area to be scanned in order to detect a certain end position of the part is the information of the other end position of the part detected so far and the known outline information of the part. Since it is determined based on, it is possible to gradually narrow the processing area, it is possible to suppress the area area on the screen to be scanned smaller than that of the conventional method, as a result,
The image processing time is shortened and the position detection speed of the component is increased.

[0008]

An embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the structure of a component detecting apparatus according to the present invention. In this embodiment, the position of a minute electronic component having a plurality of pins around it is detected.

First, the structure of the component detecting device will be described with reference to FIG. 1. The component detecting device includes an image pickup means 1 such as a television camera for picking up an image of a component which is an object, and the image pickup means 1.
Storage means 2 for storing the image obtained by the above, and edge detection means 3 for detecting the edge (pin edge) of the component by scanning the image stored in the storage means 2 within a designated area on the screen. And the area on the screen to be detected next by the edge detecting means 3 based on the position information of the edge (pin edge) of the component detected by the edge detecting means 3 and the known outline information of the component. The processing area determination means 4 and the positioning output means 5 are included in the processing area determination means 4.

The positioning output means 5 detects the pin row by scanning the image based on the position information of the end (pin edge) of the component obtained by the end detection means 3, and a pin row detection means 6; The calculating means 7 calculates the position (center position) of the component based on the position information of the pin row obtained by the pin row detecting means 6, and the output means 8 that outputs the result calculated by the calculating means 7. It is configured.

Here, the procedure for detecting the position of the electronic component will be specifically described with reference to FIGS. 2 is a diagram showing a processing screen, FIGS. 3 and 4 are explanatory diagrams showing a pin edge detecting method, FIG. 5 is an explanatory diagram showing a pin row detecting method, and FIG. 6 is a P portion (image density) of FIG. Enlarged view of distribution), Figure 7
Is an explanatory view showing a method of calculating the center position of a component, and FIG. 8 is a flowchart showing a position detection procedure.

The component imaged by the image pickup means 1 is displayed as W on the processing screen shown in FIG. 2, and a plurality of pins 9 ...

First, the part image W displayed on the processing screen.
The pin 9A in the upper left part is detected, and the processing area S1 of FIG. 2 is set at the time of this detection (step 1 of FIG. 8), and the pin 9A by the edge detecting means 3 is detected in this processing area S1. Detection is performed. The processing area S1 is determined by the distance range within which a component can be moved within a tray (not shown) in which the component is stored.

Then, in the processing area S1, as shown in FIG. 3, the edge detecting means 3 scans rightward every four lines from the point a at a predetermined interval (step 2 in FIG. 8). In order to find the first point (the point b in the example shown in FIG. 3) at which the image density is above a certain threshold Th1 and the width is within a predetermined range in each scanning line L0, L4, L8, L12. The scan is performed. Then, the detection point is used as the primary candidate point of the upper left pin edge, and the next secondary detection is performed.

In the secondary detection, as shown in FIG.
Only three lines are returned from the scanning line L12 passing through the primary candidate point (point b), and four lines are continuously scanned for each line from the scanning line L9. For example, in four-line continuous scanning from scanning lines L9 to L12 in FIG.
Since the positions of the pins (9C, 9D, 9E, 9F in the illustrated example) change for each of the lines L9, L10, L11, these pins 9 ... (9C, 9D, 9E, 9F) are to be obtained from the upper left pin 9A. Is not recognized. In this case, returning to the primary detection, the candidate point is searched again.

Thus, as shown in FIG. 4, the scanning line L
In four-line continuous scanning of 13 to L16, the positions A, A ', A''of the pin 9A are within a predetermined width (for example, within ± 1 pixel range) in the three scanning lines L14, L15, L16. , The pin 9A is recognized as the upper left portion pin, and the point A on the uppermost scanning line L14 is set as the pin edge point to be detected.

When the edge point A of the upper left pin 9A is detected in the processing area S1 as described above, the processing area determining means 4 determines the position of this edge point A (x1, y1) and the component. In order to detect the upper right pin 9B by the edge detecting means 3 based on the known external shape information (in this embodiment, the distance d between the pins 9 on both sides on the same side of the component). The processing area S2 to be scanned is determined (step 4 in FIG. 8). The edge point B (x2, y2) of the pin 9B is represented by the following equation expressed by the maximum inclination angle θ of the component and the distance d:

[0019]

Since it is estimated that x1 + d · cos θ ≦ x2 ≦ x1 + d y1−d · sin θ ≦ y2 ≦ y1 + d · sin θ, the area of the processing area S2 including the edge point B as shown in FIG. It is smaller than that of the processing area S1.

Thus, the processing area S by the edge detecting means 3
The edge point B of the pin 9B in the upper right portion is detected by the leftward scanning in step 2 (step 5 in FIG. 8) as described above (step 6 in FIG. 8). Since the area is smaller than that of the processing region S1, the image processing time is shortened and the detection speed of the edge point B is increased.

When the edge points A and B of the pins 9A and 9B are obtained in this way, the points A and B are obtained as shown in FIG.
Scanning is performed by using a straight line L obtained by offsetting a straight line connecting the lines by a predetermined amount as a scanning line, and an image density distribution along the scanning line L as shown in FIG. 8 is obtained by this scanning (step 7 in FIG. 8). Also, a plurality of straight lines L (for example, 3
In the case of a book, one line is provided before and after the straight line L), and the image density distributions on the straight line L are added together to further improve the accuracy.

Next, the edge point C of the lower left pin 9C and the edge point D of the lower right pin 9D (see FIG. 2) are detected by the same procedure as described above (steps 8 to 13 in FIG. 8). ,
The image density distribution between the left and right pins 9C and 9D is obtained (FIG. 8).
Step 14). At this time, the end portion detecting means 3 moves to the pin 9
Processing areas S3, S to be scanned to detect C, 9D
4 (see FIG. 2) is set by the processing area determining means 4 (steps 8 and 11 in FIG. 8), but the processing area S3 is based on the already obtained position information of the edge points A and B and the external shape information of the parts. And the processing area S4 is determined by the edge points A, B,
Since it is determined based on the position information of C and the outer shape information of the parts, the areas of these processing regions S3, S4 become gradually smaller as shown in FIG. 2, and the processing regions S1, S2, S3, S4.
Area decreases in this order. Therefore, the edge point A,
The image processing time required for detecting B, C, and D is gradually shortened, and the detection speed is gradually increased.

As described above, the left and right end pins 9 on the upper and lower sides
When the image density distributions between A and 9B and between 9C and 9D are obtained, the pin row detecting means 6 detects the positions of the pin rows 9 on the upper and lower sides based on this image density distribution (step of FIG. 8). 15), the calculating means 7 calculates the centers of the pin rows 9 ... on the upper and lower sides based on the position information of the pin rows 9 ... Detected by the pin row detecting means 6 (step 1 in FIG. 8).
6).

Here, the method of detecting the positions of the pin rows 9 on the upper side will be described. Four points g, h, i and j points across a certain threshold Th2 in the image density distribution shown in FIG. 6 were obtained. After that, the left and right edge points m1 and m2 are obtained by interpolation,
Let these midpoints m be the midpoints of the pins 9. Then, if the center of gravity of the midpoint coordinates of all the pins 9 obtained in this way is calculated, that point is the midpoint M of the pin row 9 on the upper side.
1 (see FIGS. 5 and 7).

Similarly, the middle point M of the lower row of pin rows 9 ...
2 (see FIG. 7) is required.

By the way, the procedure for obtaining the midpoints M1, M2 of the pin rows 9 on the upper and lower sides has been described above, but the midpoints M3, M4 of the pin rows 9 on the left and right sides (see FIG. 7). Is obtained in exactly the same manner (steps 17 and 18 in FIG. 8). That is, the edge points E, F, G, H (see FIGS. 2 and 7) in the pin rows 9 on the left and right sides are obtained by the same procedure as steps 1 to 14 in FIG. However, the processing area for detecting the edge points E, F, G, and H is gradually narrowed for the same reason as described above, so that the image processing time is further shortened and the detection can be performed at high speed. Will be able to. Also, the edge point A,
Edge points E, F, according to the outline information of B, C, D and parts
It is also possible to obtain the image density distribution without detecting G and H. By doing so, the image processing can be further speeded up.

When the middle points M2, M2, M3 and M4 of the upper and lower and left and right pin rows 9 are obtained, as shown in FIG. 7, the opposite middle points M1 and M2, M3 and M4. The center of gravity G of the whole part is calculated by finding the intersection of the straight lines connecting the two parts, and the position and inclination of the part are calculated (step 19 in FIG. 8). The result is output by the output means 8 of the positioning output means 5. It is output, and the electronic component is positioned by the output result.

[0028]

As is apparent from the above description, according to the present invention, the processing area to be scanned for detecting one end position of a part is the other end part of the part which has been detected so far. Since it is determined based on the position information and the external information of known parts, the processing area can be gradually narrowed, and the area of the screen area to be scanned can be kept smaller than that of the conventional method. Therefore, it is possible to obtain the effect that the image processing time can be shortened and the component position detection speed can be increased.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a component detection device according to the present invention.

FIG. 2 is a diagram showing a processing screen.

FIG. 3 is an explanatory diagram showing a pin edge detection method.

FIG. 4 is an explanatory diagram showing a pin edge detection method.

FIG. 5 is an explanatory diagram showing a method of detecting a pin row.

6 is an enlarged view of part P (image density distribution) of FIG.

FIG. 7 is an explanatory diagram showing a method of calculating the center position of a component.

FIG. 8 is a flowchart showing a position detection procedure.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Imaging means 2 Storage means 3 Edge detection means 4 Processing area determination means 5 Positioning output means 6 Pin row detection means 7 Calculation means 8 Output means 9 pins A to H Edge points (ends) of parts S1 to S4 Processing areas W Parts (image)

Claims (2)

(57) [Claims] 1. An area including at least one end of a part is designated on a processing screen, the position of one end of the part is detected by scanning an image in the area, and position information of the one end of the part is detected. Area that includes the other end of the part is determined based on the external information of the part and the part, and the position of the other end of the part is detected by scanning the image in the area. A component detecting method characterized in that a region including the next one end of a component is determined based on the position information of the end and the outer shape information of the component, and thereafter all the end positions of the component are detected in the same manner. 2. An image pickup means for picking up an image of a component, an edge detection means for detecting an edge of the component by scanning an image obtained by the image pickup means within a designated area, and the edge detection means. The processing area determining means determines the area on the screen to be detected next by the edge detecting means based on the detected positional information of the edge of the component and the external shape information of the component, and the edge detecting means. A component detecting device comprising a positioning output means for calculating the position of a component by scanning an image based on the position information of the end of the component and outputting the result.