JP3079640B2 - Contour detection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method in a visual recognition device for detecting the position of an electronic component in a non-contact manner at a high speed in an electronic component mounting apparatus. And a method for detecting an outline of an outer shape of an image obtained by the method.

[0002]

2. Description of the Related Art In recent years, in the field of electronic component mounting, a technique for mounting electronic components on a circuit board at high speed and accurately has been required. For this reason, digital image processing technology that captures electronic components with a video camera or the like, processes image signals obtained by digitizing the video signals at high speed, and performs non-contact and high-speed position detection of electronic components tends to be used. It is in. Conventionally,
As a position detection method utilizing digital image processing technology, a boundary point where the image pattern is in contact with the outside is detected to detect a contour line that is a series of the boundary points, and the center of gravity of the figure formed by the contour line is electronically determined. A method of detecting the position of the center of a component has been used. A conventional method for detecting a boundary point between image patterns will be described with reference to FIG.
In FIG. 5, reference numeral 50 denotes an image pattern of an electronic component, 51 denotes a hole in the image pattern, and 52 denotes a rotation center of a nozzle 53 which is sucking the electronic component at a starting point of a boundary point search. FIG.
In (a), the search is performed in the left direction 54 starting from the start point, and the boundary point 55 is detected. And from the boundary point 55,
Contour lines 56 are obtained by detecting adjacent boundary points one after another. Next, it is determined whether or not the obtained outline 56 is an outline corresponding to the outer shape of the image pattern, by determining that the number of boundary points of pixels constituting the outline is larger than a predetermined number Nd of boundary points. Depending on whether or not. If the number of boundary points is larger than the predetermined number Nd of boundary points, it is determined that the outline 56 is an outline corresponding to the outer shape of the image pattern 50. Number of boundary points Nd as a reference for judgment
Is usually set to, for example, about one-fourth of the outer peripheral dimension based on the outer peripheral dimension of the processing object.

FIG. 5B illustrates a case where the number of boundary points constituting the outline 56 is smaller than a predetermined number Nd of boundary points, and thus the detection of boundary points is performed again. As the search direction at this time, a direction 57 different from the first search direction 54 is selected, a boundary point 58 is detected, adjacent boundary points are detected one after another to obtain a contour 59, and the contour 59 is formed. The number of boundary points to be performed is compared with a predetermined number Nd of boundary points. As a result, the number of boundary points constituting the contour line 59 is set to a predetermined number N of boundary points.
Since it is larger than d, it is determined that the outline 59 is a correct outline corresponding to the outer shape of the image pattern 50.

[0004]

In the above-described conventional method for detecting a contour line, if the number of boundary points forming the contour line is larger than a predetermined number Nd of boundary points, the contour line of the outer shape of the object is determined. Since it is determined that it has been detected,
For example, as shown in FIG. 5C, when the circumference of the hole 51 becomes slightly longer for some reason, such as a defect in the shape of an electronic component or a defect in illumination at the time of imaging, depending on how the boundary point number Nd is determined. This may be mistaken for an outline.

An object of the present invention is to solve the above-mentioned conventional problems and to provide a high-speed and highly reliable contour detection method.

[0006]

In order to achieve the above object, in the method of detecting a contour line according to the present invention, a point inside an image pattern of an electronic component to be processed is set as a starting point of a boundary point search as in the conventional method. A rectangle circumscribing the detected contour is created, and it is determined that the contour has been correctly detected only when the search start point is included inside the rectangle.

[0007]

According to the present invention, the contour line corresponding to the outer shape of the object to be processed can be accurately detected even for an image pattern including a cavity by the above method.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a flowchart showing an outline of an embodiment of the present invention, and FIG. 2 is a block diagram showing an outline of an embodiment of the present invention. The first step is an image pattern input step. Mounting nozzle 1 in which electronic component 10 is provided with illumination means 11
2 and is held. A video signal obtained by imaging the electronic component 10 with an imaging unit 13 such as a video camera is digitized by a video signal digitizing unit 14 and input to an image pattern storage unit 15 as an image pattern. The second step is a boundary point search direction determining step. In a step of determining a first direction for searching for a boundary of the image pattern starting from one point inside the image pattern, the central processing unit 16 sets an initial value in the boundary point search direction storage unit 17. The third step is a boundary point search step. 1 inside the image pattern
This is a step of searching for a boundary of the image pattern in the direction determined in the second step with the point as a starting point. Boundary point searching means 18
Reads the search direction from the boundary point search direction storage means 17 and searches in that direction. When the search is completed, the boundary point searching unit 18 notifies the central processing unit 16 of the position of the searched boundary point. The fourth step is a contour detection step.
With one point on the boundary of the image pattern obtained in the third step as a second starting point, the contour line detecting means 19 successively detects adjacent boundary points, and detects a contour line which is a set of boundary points. .

The fifth step is a contour right / wrong judgment step.
It is determined whether the contour detected in the fourth step is a contour corresponding to the outer shape of the object. When detecting a boundary point of the image pattern in the contour detection step of the fourth step,
The maximum and minimum coordinates (Xmin, Ymin) and (Xmax, Ymax) of the boundary point are obtained and stored in the central processing unit 16. These two points represent a rectangle circumscribing the detected contour. If the first start point of the boundary point search step of the third step is included in the rectangular area, it is determined that a correct contour line has been detected. If the first start point is not included in the rectangular area, it is determined that a correct contour has not been detected. The sixth step is a boundary point search direction changing step. If it is determined in the fifth step that a correct contour line has not been detected, the process returns to the third step by changing the boundary point detection direction. Only the search direction is changed without changing the first starting point of the third step. The center position and the inclination of the image pattern are further calculated from the outline of the image pattern obtained by the above processing, and the center position and the inclination are sent to the host controller 20.

Next, as a specific embodiment of the present invention, a method for detecting a contour line corresponding to the outer shape of a perforated electronic component will be described with reference to FIGS. In FIG. 3, 2
1 is an image pattern of a perforated electronic component, 22 is a hole, and 23 is a rotation center of the nozzle 12 which is sucking the electronic component at a first starting point of the search process. In FIG. 3A, the search is performed with the search direction 24 set to the left, and one point 25 on the boundary of the image pattern 21 is detected. Next, in FIG.
, The boundary points adjacent to each other are successively detected while proceeding in the direction 26 in which the image pattern is viewed rightward, and a contour line, which is a set of boundary points, is detected. FIG. 3C shows a state in which the outline detection has been completed. Since the outline detection started from the point 25 has returned to the point 25, the outline detection processing is completed. Here, the maximum / minimum coordinates (Xm
in, Ymin) 28 and (Xmax, Ymax) 2
9 are determined at the same time, and the coordinates of these two points represent a rectangle 27 circumscribing the contour. In FIG. 3C, since the first starting point 23 of the search step is not included in the rectangular area, the central processing unit 16 determines that the correct contour has not been detected, and proceeds to the retry processing. FIG. 4 illustrates the retry processing. 4, the same parts as those in FIG. 3 are denoted by the same reference numerals, and description thereof will be omitted.

In FIG. 4 (a), the search direction is first turned. The search direction 24 in FIG.
Point 31 on the boundary of is detected.

Next, in FIG. 4B, adjacent boundary points are detected one after another while proceeding in the direction 32 in which the image pattern is viewed rightward from the point 31 as a starting point, and a contour line which is a set of boundary points is detected. .

FIG. 4C shows a state in which the detection of the outline has been completed. Since the outline detection started from the point 31 has returned to the point 31, the outline detection processing is completed. Here, the maximum and minimum coordinates (Xmin, Ym) of the boundary point
in) 34 and (Xmax, Ymax) 35 are determined at the same time, and the coordinates of these two points represent a rectangle 33 circumscribing the contour. In FIG. 4C, the starting point 23 of the search step is included in the rectangle 33, and it can be seen that the correct contour is detected by the retry processing. If the starting point 23 is not included in the rectangle 33, the search direction is changed and retries are repeated until a correct contour is detected.

[0014]

As is apparent from the above description, according to the contour detection method of the present invention, it is possible to accurately detect the contour of the object without being affected by the cavity in the image pattern. Therefore, the possibility of being affected by the shape of the electronic component or a defect in the illumination is reduced, and more reliable contour detection and, therefore, more reliable position detection can be performed.

[Brief description of the drawings]

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

FIG. 2 is a block diagram showing an outline of an embodiment of the present invention.

FIG. 3 is a diagram illustrating a contour detection method according to the present invention.

FIG. 4 is a view for explaining a contour detection method according to the present invention;

FIG. 5 is a diagram illustrating a conventional contour detection method.

Continuation of front page (72) Inventor Takashi Shimizu 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (56) References JP-A-63-101979 (JP, A) JP-A-60-237582 (JP) , A) JP-A-2-110777 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06T 1/00 G06T ⁷ /00-7/60 G06T 9/20 G06T 11/80

Claims (1)

(57) [Claims] 1. A first step of obtaining a digitized image pattern of a recognition target, a second step of determining a search direction in a next third step, and the search from a predetermined first start point in the image pattern. Search in the direction to find 1
A third step of detecting points, a fourth step of performing a search with one point on the contour obtained in the third step as a start point, and obtaining a contour around the contour, and a contour obtained in the fourth step. A fifth step of making a rectangle circumscribing the line and determining that the contour is not correct if the first starting point is within the rectangle; otherwise, determining that the contour is not correct; A step of changing the direction of the search when the contour is determined to be incorrect in the step and returning the processing to the third step.