JPS5991584A - Method for discriminating correctness/incorrectness of circular pattern - Google Patents

Abstract

PURPOSE:To shorten processing time and to reduce the production cost of a circular pattern by calculating an external defective area permissible circle and an internal defective area permissible circle from the positional coordinates of the edges of a regular circle caught by a scanning line and by comparing said areas with the coordinates of the edge or the regular circle. CONSTITUTION:The center coordinates O of a circular pattern 1 are calculated from the X coordinates P1, Q1 of the edges measured by an optical scanning line 3-1 and a radius R. The external defective area permissible circle 4 and the internal defective area permissible circle 5 are calculated from external fatal defect permissible width E and internal fatal defect permissible width F, and whether the positional information of edges measured by the 2nd scanning line 3-2 or after of the circular pattern 1 exists in a fatal defect permissible area 6 of a space between the circles 4, 5 or not is discriminated to determine the existence of a defect.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for determining the acceptability of circular patterns, and more particularly to a method for optically detecting defects.

Conventionally, in the field of image processing, the method of detecting defects such as circular pattern protrusions, defective pinholes, etc. is to scan the inspected object with a -dimensional optical scanner or two-dimensional optical scanner, etc.
A method is employed in which image information is temporarily stored in a storage device in units of picture elements, and defects are detected by comparing each picture element with images of manufacturing design information. However, this conventional method does not require a huge storage device, and since the processing is done on a pixel-by-pixel basis, it takes time to judge defects, and man-hours to correct errors in alignment. I am one who has faced difficulties that I cannot do.

An object of the invention is to provide a method for determining the acceptability of a circular pattern, which solves the above-mentioned conventional problems.

According to the present invention, when detecting a fatal defect occurring in a prescribed circle with a radius R, one-dimensional light scanning is performed from outside the prescribed circle at a fixed pitch, and the defect is captured by a scanning line that first enters inside the prescribed circle. Measure the position coordinates PI and P2 of the edges of the prescribed circle, find the center position coordinates within the regulations from the position coordinates PI and B2, and create an external defect area tolerance circle and an internal defect area tolerance circle around the center position coordinates. A method for determining the acceptability of a circular pattern is provided, which is characterized in that it determines the presence or absence of the positional coordinates of an edge within the specified range to be measured in the next and subsequent scans within the range sandwiched between the two tolerance circles. can get.

Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings.

1 to 3 show a state in which a circular pattern 1 is captured for the first time in a field of view 2 of a -dimensional optical scanner. Since the position of the circular pattern 1, which is the object to be inspected, is unknown, it is not known in which state among FIGS. 1 to 3 the image of the circular pattern 1 entering the field of view 2 of the scanner will be captured. Therefore, if the conditions shown in Figures 1 to 3 are not taken into consideration, errors will occur in the calculated values of the external defect area tolerance circle 4 and the internal defect area tolerance circle 5 as shown in Figure 4, resulting in problems. .

FIG. 5 explains a method for determining the quality of the circular pattern 1, taking into consideration the removal of the above-mentioned defects and the conventional problems. It is assumed that the radius R, the external critical defect tolerance width E, and the internal critical defect tolerance width F of the circular pattern 1 are known from design information and other sources.

X position of the edge of circular pattern 1 measured by the first optical scanning line 3-1 in circular pattern 1 Each coordinate is PI
, Ql, the center position coordinate O of the circular pattern 1, which removes calculation errors occurring in the states shown in FIGS. 1 to 3, can be calculated using the following equation.

X coordinate of the center position coordinate 0 - (Ql - PI>/2 Furthermore, since the center position coordinate data of the circular pattern 1 was obtained, the X coordinate on the circumference of the external defect area tolerance circle 4 is Al, A2
, A3. , , , , An-1, , , , , B
l, B2°B3. ......Bn-! ,...
... and the X coordinate c1. on the circumference of the internal defect area tolerance circle 5. c.
2. ......, Cn-2,..., Dl.

B2. . . . , Dn-2, . . . can also be calculated using the following formula.

An's X coordinate - 0's X coordinate - An's Y coordinate ff = o's Y coordinate - nK Bn's X coordinate = 0's X coordinate frame Bn's Y coordinate - 0 (7) Y coordinate - nK Cn's X coordinate = 0's X coordinate Cncr) Y coordinate - 0's Y coordinate - (n K+E ) Dn
X coordinate of - X coordinate of 0 - Y coordinate of Dn - Y coordinate of 0 - (n K-1-E ) However, imaginary results are not treated as data.

The quality determination of the circular pattern is based on the edge position information measured after the second optical scanning line 3-2 of the circular pattern 1.
The presence or absence of a defect is determined by determining whether the defect exists in a fatal defect tolerance area 6, which is a space narrowed between an external defect area tolerance circle 4 and an internal defect area tolerance circle 5.

That is, the X position coordinate B2 of the edge measured by the second optical scanning line 3-2. Q2 is the coordinate A obtained by the above calculation
It is determined whether or not it exists in the range from 1 to B1. (In this case, Cn and Dn have no X coordinate value that matches the Y coordinate value.) Similarly, A2 <B3<CI DI<Q3<B2, and there is no edge information in the range 01 to D1 (An
, Bn, Cn, Dn (there exists an X coordinate value with which the value of the D Y m marker matches) P4. Similar operations are performed for Q4 and thereafter until the information on the edge of the circular pattern 1 to be protected disappears. 6 to 11 are examples of determination of a circular pattern having a fatal defect by the method of the present invention. FIG. 6 shows a pattern in which the circular pattern has a defect, and measurement edge data exists in the pronation and pronation of Cn to Dn shown in FIG.

In Figure 7, the circular pattern has a protrusion, and the measured etch data exceeds the calculated data by 3'' in the Y-axis direction. In Figure 8, the circular pattern has a defect. Sha in the pattern, All and C in the middle of measurement
Edge data disappears from the range of n, and Cn and Dn
Edge data exists within the range. In FIG. 9, the circular pattern has protrusions, and edge data disappears from the range of An and Cn during measurement. FIG. 10 shows a pattern with pinholes, and all edge information exists within the critical defect tolerance region 6 shown in FIG. 5, and Cn
Edge data exists within the range of and Dn. FIG. 11 shows the case W, where an island exists near the circular pattern. In this case, an external defect area tolerance circle 4-1 and an internal defect area tolerance circle 5-1 are newly assumed in the island portion.

In any case in FIGS. 6 to 11 above, it is determined to be defective.

As described above, the present invention uses the radius of the inspection circular pattern, the internal fatal defect tolerance width, and the external fatal defect tolerance 1 as the inspection parameters.
It is possible to discover a fatal flaw in a circular pattern at any position f6 by simply giving tsu1, and since only the etch data is handled, the pass/fail judgment method can be made at a lower cost and in a shorter processing time. Realized and practical price is a dog.

4. Figures 1 to 3 are diagrams showing the -dimensional optical scanning and the initial state fluctuation of the inspected bag, and Figure 4 shows the external defect area tolerance date and internal defect area allowance. FIG. 5 shows the algorithm of the present invention, and FIGS. 6 to 11 are diagrams showing the determination of a circular pattern having a flame defect according to the '2# invention.

Reference numeral 1 in the figure: circular pattern, 2: field of view of the scanner, 3: scanning line 1.4: allowable external defect area. , 4-1... Assumed external defect area tolerance circle, 5... Internal defect area tolerance circle, 5
-1... Assumed internal defect area tolerance circle, 6...
...The tolerance for fatal defects will increase.

Claims (1)

[Claims] When detecting a fatal defect occurring in a specified circle with radius R, -dimensional optical scanning is performed at a fixed pitch from outside the specified circle, and the etch of the specified circle captured by the scanning line that first enters inside the specified circle is detected. The position coordinates PI, P2 are opened, and the position coordinates P
Find the center position coordinates of the regulation circle from L and P2, calculate the external defect area tolerance circle and the internal defect area tolerance circle around the center position coordinates, and calculate the next time and subsequent times within the range between both tolerance circles. A method for determining the acceptability of a circular pattern, characterized by determining the presence or absence of the positional coordinates of the edge of a prescribed circle measured by scanning.