JPS63269001A - Method for recognizing body - Google Patents

Abstract

PURPOSE:To recognize the position of a body with high accuracy, by extracting the predetermined area of the image of the body irradiated wit slit beams crossing at a right angle to detect the shift between the center and a reference position and receiving said shift within predetermined accuracy before calculating the shift quantity of rotation. CONSTITUTION:The mount reference position O0 of the chip 2 in an image wherein the center coordinates of an image pickup visual field are (Ix, Iy) is irradiated with slit beams LX, LY crossing at a right angle. The picked-up image is converted to a digital value by an A/D converter to be stored and binarized on the basis of a predetermined value to obtain an image wherein only the irradiated parts of the slit beams LX, LY are extracted. Windows WX, WY are set from said image. The sizes of the windows are set from the size of the chip 2 and the widths of the slit beams LX, LY and the peripheral distribution in respective X- and Y-directions is calculated with respect to the windows WX, WY. The end edge of the upper surface of the chip 2 is calculated on the basis of a threshold value S from the analytical result and the center O1(XC, YC) of the chip is calculated. The shift quantity from the mount position O0 is calculated to move the slit beams and, when the shift quantity is within a predetermined valve, an angle theta of inclination is calculated. By this constitution, the position of a body having a complicated shape can be recognized with high accuracy.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to an object recognition device that recognizes the position, rotational deviation, etc. of chip components, etc.

(Prior Art) Electrical devices such as home appliances and OA devices are becoming smaller. Along with this, the circuit boards used in these home appliances are also becoming smaller. For example, instead of circuit boards with lead wires, circuit boards that are made into chips and can be equipped with small resistance capacitors, etc. are being used. There is. Currently, there is a demand for high-density mounting on circuit boards by miniaturizing such mounted components.

To attach such chip components to a circuit board, a mounting device automatically temporarily attaches the chip components to a circuit board on which a predetermined wiring pattern has been formed, and soldering is performed using a device, etc. finish.

However, when attached to a circuit board, they often fall off or become misaligned. If this is detected after soldering, a great deal of effort must be spent to correct it. For this reason, it has been necessary to detect chip components of such circuit boards for falling off, misalignment, etc. before soldering. This work was performed visually by a worker before soldering.

In response, technology is being developed to automate the inspection of circuit boards.

(Problems to be Solved by the Invention) Conventional visual inspection by workers requires a large number of personnel. Furthermore, since multiple chip components are mounted on a complex circuit board, there are many inspection errors, and the work involves performing such detailed inspection work all day long, which causes eye fatigue and poses health management problems. was occurring.

In contrast, the circuit board is imaged with an imaging device.

Techniques are being developed to process the image data obtained in order to detect the falling off, misalignment, etc. of chip components.

However, with such technical means, incorrect judgments are often made because the target chip components, etc. are searched for or extracted from a mixture of wiring patterns, printed characters, and multicolored and multishaped components on the circuit board. arise. Therefore, it is still insufficient in terms of inspection accuracy and reliability, and especially when the target chip component has a shape other than a normal rectangular parallelepiped, the inspection accuracy and reliability are insufficient.

Therefore, in order to solve the above-mentioned problems, the present invention aims to recognize the position of a target object with high accuracy when automating the inspection of circuit boards, without being affected by the complex shape, its color, or the shape of the target object. The purpose is to provide an object recognition device that can recognize objects.

[Structure of the Invention] (Means for Solving the Problems) Two slit lights that are orthogonal to the area containing the object to be recognized at a predetermined reference position and movable at an angle within a predetermined range. irradiate. The area including the object to be recognized, which is irradiated with these two orthogonal slit lights, is imaged from directly above. After digitizing and storing the captured image information, only a predetermined area including the two slit lights on the upper surface of the object to be recognized is extracted. This extracted image information is binarized, and marginal distributions of the binarized image information are taken in the X and Y directions. From this result, the center position of the object to be recognized is determined, and then the amount of deviation between the center position and the reference position is determined. This deviation is compared with a predetermined accuracy, and if it is not within the predetermined accuracy, the center position of the detected object to be recognized is set as a new reference position, and the irradiation reference position of the two slit lights is corrected. Furthermore, if the amount of deviation is within a predetermined standard, the rotational deviation of the object to be recognized is determined.

(Operation) By providing the above steps, the recognition object can be detected from the peripheral distribution in two directions of The center position and the amount of deviation between the center position and the reference position are detected. Compare this with a predetermined accuracy. If the result is not within the predetermined accuracy, the reference position for irradiating the slit light is set and repeated until the result is within the predetermined accuracy. Furthermore, if the result is within a predetermined accuracy, the amount of rotational deviation is calculated and object recognition information is obtained.

(Example) An example of an object recognition device using the method of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an object recognition device. Circuit board (1
) Two slit illumination devices (3a) (3b) are arranged diagonally above the top surface of the chip component (2) mounted in (4b) is a slit illumination device (
The galvano mirrors (4a) and (4b) can be controlled by the galvano mirror control circuit (5). It is set in. Further, an imaging device (6) is arranged in a direction perpendicular to the circuit board (1), and an A/D converter (7) is connected so that the image signal outputted by the imaging device (6) can be digitized.
It is connected so that the results can be stored in an image storage memory (8). Furthermore, a binarization circuit (9) is connected so that the data stored in the image storage memory (8) can be retrieved, and the following sequentially includes a window setting circuit (10) and a peripheral distribution detection circuit (
11), a center position and deviation amount detection circuit (12), a comparison circuit (13), and a rotational deviation amount detection circuit (14) so that the output results can be sent thereto. Further, a control circuit (15) capable of controlling each of these circuits and mechanisms is connected to each of them.

Next, the operation will be explained using an apparatus according to an embodiment of the method of the present invention. FIG. 3 shows a flowchart of the method of the invention. First, the chip component (2) in the image captured by the imaging device (6) whose center coordinates of the imaging field of view are (Ix, Iy)
Mount reference position. The control circuit (15) sends rotation angle data to the galvano mirror control circuit (5) so as to irradiate orthogonal slit light (LX) (LY) to 0 (Ox, Oy), and the galvano mirror control circuit (5) Drive the galvano mirrors (3a) and (3b).

At this time, each mirror rotation angle data (MXD, MYD) is the scanning resolution (Mx, MYD) of the galvano mirror in the X direction and Y direction.
My), height dimension (I+) of chip component (2), thickness (T) of circuit board (1), galvanometer mirror (la) (
3b) height data (11) at each scanning position,
The correction data (Kx, Ky) of (T) is obtained using the following formula.

MXD= (OX-IX)/MX+lI+T+KXMY
D-(Oy-1y)/My+H+T+Ky The slit light (LX) (LY) irradiated in this way is imaged by the imaging device (6), passed through the A/D converter (7), digitized, and stored. Store in memory (8). The data stored in this storage memory (8) is transferred to the binarization circuit (
9) is read and binarized with a predetermined value so that only the portion illuminated by the slit light (LX) (LY) can be extracted, thereby obtaining an image as shown in FIG. 2(a). Window setting circuit (10) from the image obtained in this way
is the window for confirming the X-direction position and Y-direction position (Wx
) (My).

The size of the window at this time is set based on the size of the chip component (2) and the width of the slit light (t, x) (LY). Then, the peripheral distribution detection circuit (11) is shown in Fig. 2 (
b) As shown in (c), the marginal distribution in the X direction with respect to the window (Wx) and the marginal distribution in the Y direction with respect to the window (wy)
Find the marginal distribution of each direction. From the peripheral distribution results in these two directions, the edges (XL,
R, YL, YR) are determined using the threshold value S, and the center position of the chip component is 0. (Xc, Yc) is determined from this edge data using the following formula.

Xc = (XL + XR) / 2 Yc = (YL + YR) / 2 The chip component center position (Xc, Yc) was determined in this way. This is fine if the chip component (2) is not tilted, but if it is tilted If so, the center position of the chip component (2) may differ from the correct center position (0). Therefore, the center position and deviation amount detection circuit (12) detects the initially set mounting reference position of the chip component (2). Calculate the amount of deviation from 0 (Ox, Oy). The comparison circuit (13) to which the results obtained by the center position and deviation amount detection circuit (12) are sent compares the mount reference position 0 (Xc, Yc) and the chip component center position 0.
(Xc, Yc) is larger than a predetermined amount n+1 d, a signal is sent to the control circuit (15), which controls the mount reference position 0
), set (Xc, Yc) as a new mount reference position 01 (Xc, Yc), and insert two slit lights on the top surface of the chip component (2) of this new mount M level lid 0□ (Xc, Yc). Send rotation angle data to the galvano mirror control circuit (5) so that (LX) and (LY) are orthogonal, drive the galvano mirrors (3a) and (3b), and irradiate two slit lights (LX) and (LY). . The above-described operations are performed in response to this. In addition, the comparison circuit (13) compares the mount reference position On (Xc, Yc) and the chip component center position,
When the difference from 0 (xc, Yc) is within a predetermined ff1d, the data of the edge (YL, YR) in the Y direction is sent to the rotational deviation total calculation circuit (14), and it is finally detected from the peripheral distribution in the Y direction. From the edges (YL, YR) in the Y direction, the tilt angle θ of the chip component (2) is calculated using the following formula.

In addition, (Ix, Iy) and galvano-mirror (3a) (3b)
The origins of must match. In other words, when the galvano mirror (la) (3b) is at the origin, the slit light (LX) (LY) is at the center coordinates (lx, Iy) of the imaging field of view.
are perpendicular to each other. In the above embodiment, calculation was made to irradiate the slit light onto the top surface of the object to be recognized, but if the inclination of the object is large, the irradiated light may not intersect on the top surface of the object. This is to avoid complicate calculations in such cases, and if the inclination of the object to be recognized is small, the irradiation position may simply be set on the substrate. Further, although the rotational deviation amount is calculated from the last edge data in the Y direction, this is not limited to the last data, and the edge data to be calculated may also be edge data in the X direction.

[Effects of the Invention] As described above, the object recognition device according to the method of the present invention can recognize the position of an object in two directions by simultaneously irradiating the slit light in two directions, X and Y, and can recognize the object in two directions. The position and angular shift of an object can be recognized regardless of its shape. Furthermore, since the recognition is based on the height information of the chip component, the target component can be recognized without being affected by the wiring pattern, printed characters, resist, etc. on the circuit board, or the color of the component. Furthermore, for parts within the same field of view, the XY slit light can be irradiated and scanned arbitrarily to the parts to be recognized, thereby speeding up the inspection.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of an object recognition device according to the method of the present invention, FIG. 2 is a state diagram showing an image taken by the object recognition device and its processing state, and FIG. 3 is a flowchart also used in this example. DESCRIPTION OF SYMBOLS 1... Circuit board, 2... Chip components, 3a, 3b.
...Slit illumination device, 4a, 4b... Galvano mirror, 5... Galvano mirror control circuit, 6... Imaging device, 7... A/D converter, 8... Image storage memory, 9 ...Binarization circuit, 10...Window setting circuit,

Claims (2)

[Claims] (1) An initial illumination step of irradiating an area containing the object to be recognized with two slit lights having an angle within a predetermined range and orthogonal to a predetermined reference position where the center of the object to be recognized should be; An imaging step in which the area including the object to be recognized, which has been irradiated with two orthogonal slit lights in this initial illumination step, is imaged from directly above, and an image information storage step in which the image information captured in this imaging step is digitized and stored. and a slit light image information extraction step of extracting only a predetermined size area including two orthogonal slit lights irradiated onto the object to be recognized from this image information storage step, and this slit light image information An image information binarization process that binarizes the image information extracted from the extraction process, and a marginal distribution detection that takes the marginal distribution in the X direction and Y direction of the image information binarized by this image information binarization process. a center position detection step for detecting the center position of the object to be recognized from the results of this peripheral distribution detection step; and a positional deviation amount for detecting the amount of deviation between the center position detected in this center position detection step and the above reference position. a detection step, a comparison step of comparing the amount of deviation detected in this positional deviation amount detection step with a predetermined accuracy, and the center position detection described above until the accuracy is within the predetermined accuracy based on the comparison result in this comparison step. An object characterized by having an illumination reference position correction step of correcting the center position of the object to be recognized detected in the process to a new reference position, and a rotational deviation amount calculation step of calculating the amount of rotational deviation after the completion of the comparison step. How to recognize. (2) The object recognition method according to claim 1, wherein in the illumination step, the slit light can be moved with respect to a plurality of objects to be recognized.