JPH10148516A - Image inspection apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an image inspection apparatus, by which the accuracy of the judgment of a quality can be enhanced, by extracting, when the interval between adjacent candidate points is at a threshold value or lower, only the candidate point on one side, which satisfies a definite condition and judging the quality of an object to be inspected. SOLUTION: An image which is picked up by a video camera 1 is converted by an A/D conversion part 2 so as to be stored in an image memory 3, and it is reconverted by a D/A conversion part 4 so as to be displayed on a monitor 5. In a search processing part 12, the imaged image which is stored in the image memory 3 is searched by using a model image which is stored in a model memory 9, and coordinate data on an agreement point is stored in a data memory as coordinate data. A search-candidate thinning-out processing part 13 extracts only a candidate point, on one side, which satisfies a prescribed condition when the interval between adjacent candidate points is at a threshold value or lower, the candidate point is stored in the data memory 10 as coordinate data after a thinning-out processing operation, and a candidate point on the other side is thinned out. A CPU 6 judges the quality of the lead terminal of an IC on the basis of the coordinate data after the thinning-out processing operation so as to be output to an output part 11b.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image inspection apparatus suitable for inspecting an electronic component such as a semiconductor integrated circuit having a plurality of strip-shaped or pin-shaped terminals.

[0002]

2. Description of the Related Art Conventionally, an image of a target object is picked up using an image pickup device, and the picked-up image is searched for a known image serving as a reference, and the two images are matched to judge the quality of the target object. Inspection devices are known.

When the object to be inspected is a semiconductor integrated circuit and its lead terminals are to be bent or floated, a video camera is used to image the semiconductor integrated circuit and a model representing the shape of the lead terminals registered in advance. An image is searched for a captured image, and the leading end position, pitch, and the like of the lead terminal are obtained from the position information when the two images match, and the quality is determined.

[0004]

However, since the surface of the lead terminal is flawed or the plating state is not uniform, the discrimination threshold is set to an appropriate value when finding the leading end position of the lead terminal by search processing. If there is, it can be determined that there is a lead terminal, but if the threshold value is large, the lead terminal cannot be found. The inconvenience of erroneously determining a defective product occurs.

SUMMARY OF THE INVENTION The present invention has been made to solve such a conventional problem, and an image inspection method which accurately detects the tip position of a lead terminal by a search process and does not erroneously determine a non-defective product as a defective product. It is intended to provide a device.

[0006]

According to a first aspect of the present invention, there is provided an image memory for storing an image of a target object, and a search for a target image in a reference image to determine a coincidence point between the two images. A search processing unit that detects points, and a search candidate thinning processing unit that extracts only one candidate point that satisfies a predetermined condition when the interval between adjacent candidate points is equal to or less than a threshold and thins out the other candidate point. Determining means for determining the quality of the target object based on the candidate points.

According to a second aspect of the present invention, in the first aspect, the target object is an electronic component on which a plurality of strip-shaped or pin-shaped terminals are arranged, and the search processing unit is located at a tip end position of the terminal. Is detected as a candidate point, and when the interval between adjacent candidate points is equal to or smaller than a threshold value, the search candidate thinning processing unit extracts a candidate point on the terminal side in the longitudinal direction of the terminal and thins out other candidate points.

According to the present invention, candidate points are detected by template matching in the search processing section, only candidate points satisfying a predetermined condition are extracted from the candidate points, and other candidate points are thinned out to detect the candidate points. The accuracy is increased, and the accuracy of determining whether the target object is good or bad is improved.

[0009]

FIG. 1 is a block diagram showing an embodiment of an image inspection apparatus according to the present invention. In FIG. 1, an image captured by a video camera 1 is converted into a video signal, sent to an A / D converter 2, and stored in an image memory 3 as digital image data. The image data stored in the image memory 3 is converted again into an analog video signal by the D / A converter 4 and displayed on the monitor screen 5 of the display device.

A CPU (Central Processing Unit) 6 functions as control means for controlling the entire apparatus and determining whether or not the target object is good or bad. ROM)
8, a model memory (RAM) 9 for storing model images for searching for captured images, a data memory (RAM) 10 for temporarily storing various data input from the outside, processing results by the CPU 6, and the like; An input unit 11a such as a keyboard and an output unit 11b such as a printer are connected to an input / output interface (I / O) 11 for transmitting and receiving data.

The system bus 7 has an image memory 3,
The search processing unit 12 and the search candidate thinning processing unit 13 are connected to each other. The search processing unit 12 includes the image memory 3
Is searched by the model image stored in the model memory 9 and the coordinate data of the matching point is stored in the data memory 10 as the coordinate data of the candidate point. The search candidate thinning processing unit 13 extracts only coordinate data satisfying a predetermined condition from the coordinate data of the candidate points detected by the search processing unit 12, and stores the coordinate data in the data memory 10 as the thinned coordinate data. .

Next, the operation of this embodiment will be described with reference to the flowchart shown in FIG. In the following description, among the lead terminal groups Ra, Rb, Rc, and Rd of the semiconductor integrated circuit IC shown in FIG. 3, the leading end position of each terminal of the upper lead terminal group Ra shown in the upper part of the figure is detected. An example will be described in which the inspection of the degree of bending or floating of the lead terminal is performed by the inspection.

First, the image of the semiconductor integrated circuit IC stored in the image memory 3 is searched by the model image for the upper lead terminal stored in the model memory 9.
The obtained coordinate data (SX1, SY) of the plurality of N candidate points
1) to (SXn, SYn) are stored in the data memory 10 (step S1). The coordinate data of the candidate point is, for example, the coordinates of the center of the tip of the lead terminal. This process is performed by the search processing unit 12.

Next, the obtained coordinate data (SX1, SY1) to (SXn, SYn) of the plurality of candidate points are sorted in the order of the x-coordinate position, and the sorted coordinate data is (X1, Y1).
To (Xn, Yn) in the data memory 10 (step S2).

For example, as shown in FIG. 4, the origin is set at the upper left, the x-axis is taken in the horizontal direction, the y-axis is taken in the vertical direction, and the upper lead terminals Ra1, Ra2, Ra3,. Then, points P, Q,
R, ... are detected. In addition, uneven plating W on terminal Ra1
1 as a candidate point S, and a scratch W2 on the terminal Ra2 as a candidate point T
, The uneven plating W3 on the terminal Ra3 is detected as the candidate point U.

As a result, in the coordinate data of each of the candidate points P to U, the point P is (SX1, SY1) and the point Q is (SX2, SY).
2), point R is (SX3, SY3) and point S is (SX4, S
Y4), point T is (SX5, SY5), point U is (SX6,
SY6).

When these coordinate data are sorted in the order of the x coordinate position, since S <P <Q <T <U <R, the coordinate data after sorting is such that point S is (X1, Y1) and point P is (X1, Y1). X
2, Y2), point Q is (X3, Y3), point T is (X4, Y
4) The point U is (X5, Y5) and the point R is (X6, Y6). These coordinate data are stored in the data memory 10. In the drawing, the numbers in parentheses below the points P to U indicate the order after sorting.

Next, a loop counter i is set to 2;
The terminal counter c is set to 1. In addition, 1 after thinning
The coordinate data (X1, Y1) of the point S is set as the third coordinate data (AX1, AY1) (step S3).

Next, coordinate data Xi and coordinate data AX
It is determined whether the difference from c is less than the threshold value dx, that is, Xi-AXc <dx (step S4). The threshold value dx is an appropriate value about the terminal width.

In this case, since the coordinate data Xi (i = 2) is the coordinate data X2 of the point P and the coordinate data AXc (c = 1) is the coordinate data X1 of the point S, both points S and P are the same terminal. Since it is above, “X2−X1 <dx” is satisfied, and the condition of the expression is satisfied.

Then, next, the Y coordinate data is compared to determine whether Yi <AYc (step S5). In this case, the coordinate data Y
Since i (i = 2) is the coordinate data Y2 of the point P and the coordinate data AYc (c = 1) is the coordinate data Y1 of the point S, “Y2 <Y1” is satisfied, and the condition of the equation is satisfied.

Therefore, the c-th coordinate data (AXc, AYc) after the thinning is set as coordinate data (Xi, Yi) (step S6). In this case, since the terminal counter c is 1 and the loop counter i is 2, the first coordinate data (AX1, AY1) after thinning is converted to the coordinate data (X
2, Y2).

Next, it is determined whether or not the processing has been completed for all N candidate points (step S7). If the processing has not been completed, the loop counter i is incremented to 3 (step S8), and the condition of the expression is again satisfied. Is determined (step S4).

This time, since the coordinate data Xi (i = 3) is the coordinate data X3 of the point Q and the coordinate data AXc (c = 1) is the coordinate data X2 of the point P, the two points Q and P are on different terminals. Therefore, “X3−X2> dx” is satisfied, and the condition of the expression is not satisfied.

Therefore, next, the terminal counter c is incremented to 2 and the c-th coordinate data (AX
c, AYc) as coordinate data (Xi, Yi) (step S9). In this case, since the terminal counter c is 2 and the loop counter i is 3, the second coordinate data (AX2, AY2) after thinning out is the coordinate data (X3,
Y3) is set.

Next, the processing of steps S7 to S8 is executed, the loop counter i is incremented to 4, and the condition of the equation is determined again (step S4). This time, since the coordinate data Xi (i = 4) is the coordinate data X4 of the point T and the coordinate data AXc (c = 2) is the coordinate data X3 of the point Q, since both points T and Q are on the same terminal, "X4-X3
<Dx ”, which satisfies the condition of the equation.

Then, the condition of the equation is judged again (step S5). In this case, the coordinate data Yi (i = 4) is the coordinate data Y4 of the point T, and the coordinate data AYc (c =
Since 2) is the coordinate data Y3 of the point Q, "Y4> Y
3 ", which does not satisfy the condition of the expression.

Next, the processing of steps S7 to S8 is executed, the loop counter i is incremented to 5, and the condition of the equation is judged again (step S4). This time, since the coordinate data Xi (i = 5) is the coordinate data X5 of the point U and the coordinate data AXc (c = 2) is the coordinate data X3 of the point Q, since both points U and Q are on different terminals, "X5-X
3> dx ", which does not satisfy the condition of the equation.

Then, next, the terminal counter c is incremented to 3 and the c-th coordinate data (AX
c, AYc) as coordinate data (Xi, Yi) (step S9). In this case, since the terminal counter c is 3, and the loop counter i is 5, the coordinate data (X5, X5,
Y5) is set.

Next, the processing of steps S7 to S8 is executed, the loop counter i is incremented to 6, and the condition of the equation is judged again (step S4). Since i is 6, the coordinate data Xi (i = 6) is the coordinate data X6 of the point R, and the coordinate data AXc (c = 3) is the coordinate data X5 of the point U. Since it is on the terminal, “X6−X5 <dx” is satisfied, and the condition of the expression is satisfied.

Then, the condition of the equation is judged again (step S5). In this case, the coordinate data Yi (i = 6) is the coordinate data Y6 of the point R, and the coordinate data AYc (c =
Since 3) is the coordinate data Y5 of the point U, "Y6 <Y
5 ", which satisfies the condition of the expression.

Therefore, the c-th coordinate data (AXc, AYc) after the thinning is set as coordinate data (Xi, Yi) (step S6). In this case, since the terminal counter c is 3 and the loop counter i is 6, the third coordinate data (AX3, AY3) after thinning is the coordinate data (X
6, Y6).

Thus, the processing of steps S4 to S9 is repeated, and if it is determined in step S7 that the processing of all the candidate points has been completed, the processing is terminated. As a result, the coordinate data (AX1, AY1) to (AX
m, AYm) is coordinate data (AX1, AY1)
Coordinate data of the point P, coordinate data (AX2, AY
2) The coordinate data of the point Q and the coordinate data (AX3, A
The coordinate data of the point R is stored in the data memory 10 as Y3). However, each coordinate data of the points S, T, and U is thinned out, and is not stored in the data memory 10.

The CPU 6 calculates the coordinate data (AX1, AY) of the candidate points P, Q, R,...
1), (AX2, AY2), (AX3, AY3),..., Determine the quality of the lead terminals of the semiconductor integrated circuit IC, and output the result to the output unit 11b.

In the above-described embodiment, the case has been described in which the tip position of the upper lead terminal group Ra of the semiconductor integrated circuit IC is detected. What is necessary is just to search a picked-up image by the model image for lead terminals.

However, when inspecting the left lead terminal group Rc and the right lead terminal group Rd, it is necessary to determine the pitch width between the lead terminals in the y-axis direction in step S4, and the x-axis in steps S5 and S6. It is necessary to determine the direction.

[0037]

According to the present invention, candidate points are detected by template matching in the search processing unit, and only candidate points satisfying a predetermined condition are extracted from the candidate points and other candidate points are thinned out. Increase the accuracy of candidate point detection,
The accuracy of determining the quality of the target object can be improved.

According to the present invention, when the interval between adjacent candidate points is equal to or smaller than a threshold value, that is, when the pitch width between lead terminals is equal to or smaller than the threshold value, a candidate point on the terminal side in the longitudinal direction of the terminal is extracted. Since other candidate points are thinned out, it is possible to accurately detect the position of the tip of the terminal.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a timing chart for explaining the operation of the present invention.

3A is a plan view of a semiconductor integrated circuit to be inspected, and FIG. 3B is a side sectional view thereof.

FIG. 4 is an explanatory diagram for explaining a candidate point thinning process of an upper lead terminal group of the semiconductor integrated circuit;

[Explanation of symbols]

 Reference Signs List 1 video camera 2 A / D conversion unit 3 image memory 4 D / A conversion unit 5 monitor screen 6 CPU (central processing unit) 7 system bus 8 program memory (ROM) 9 model memory (RAM) 10 data memory (RAM) 11 Input / output interface (I / O) 12 Search processing unit 13 Search candidate thinning processing unit IC Semiconductor integrated circuits Ra to Rd Lead terminal group

Claims (2)

[Claims] An image memory for storing an image of a target object; a search processing unit for searching for an image of the target object with a reference image to detect a coincidence point between the two images as a candidate point; If the interval is equal to or less than the threshold, a search candidate thinning processing unit that extracts only one candidate point that satisfies a predetermined condition and thins out the other candidate point, and determines whether the target object is good or bad based on the extracted candidate point. Means for inspecting an image. 2. The target object is an electronic component in which a plurality of band-shaped or pin-shaped terminals are arranged. The search processing unit detects a tip position of the terminal as a candidate point, and the search candidate thinning processing unit is an adjacent component. 2. The image inspection apparatus according to claim 1, wherein when the interval between the candidate points to be executed is equal to or smaller than a threshold value, a candidate point on the distal end side in the longitudinal direction of the terminal is extracted and other candidate points are thinned out.