JP2821085B2 - Inspection equipment for periodic patterns - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a periodic pattern for detecting defects such as flaws, pinholes, black spots, and dust on a test object having a periodic pattern such as a liquid crystal panel, a color filter thereof, and a shadow mask. It relates to an inspection device.

[0002]

2. Description of the Related Art FIG. 6 is a block diagram showing a configuration of a conventional periodic pattern inspection apparatus. In this inspection apparatus, the periodic pattern of the inspection object is read by the image input unit 102 for each pixel using a line sensor or the like, and the comparison unit 106 outputs an image signal output from the image input unit 102;
A signal obtained by taking a difference from the image signal obtained by delaying the image signal by one period of the periodic pattern by the period delay unit 104 and using the absolute value of the difference as a signal value (hereinafter referred to as a “difference signal”).
Get. For example, when a periodic pattern as shown in FIG. 7A is read along the line AA, an image signal as shown in FIG. 7B is obtained, and this image signal and this for one period are obtained. The difference signal from the delayed image signal (the signal shown in FIG. 7C) is as shown in FIG. 7D. The defect detection unit 108 determines that a defect exists at a location where the value of such a difference signal is equal to or greater than a predetermined threshold, and outputs a defect signal indicating the presence or absence of the defect. However, as can be seen from FIG. 7D, two peaks corresponding to one defect appear in the differential signal one cycle apart, so that it is determined which of two adjacent repeating units has a defect. There is a need to.

On the other hand, in the defect detection method disclosed in Japanese Patent Application Laid-Open No. 61-205811, each value of the difference signal is compared with a predetermined threshold value, and only a portion having a predetermined threshold value or more is compared. 1 ′ is obtained, and binarized data obtained from the first and second repeating units of three consecutive repeating units is obtained. 2 from the second and third repeat units
By taking a logical product with the quantified data, it is specified which of two adjacent repeating units has a defect. According to such a method, for example, a periodic pattern having a repetition period L as shown in FIG.
(B) Binary data of a pattern as shown in FIG.
FIG. 8 shows a logical product of the binary data.
Data of a pattern as shown in (c) (hereinafter referred to as “logical product data”) is obtained. This makes it possible to specify in which repetition unit the defect exists.
Also, by taking the logical product of the binary data,
Since the influence of a false signal or noise caused by the displacement of the sampling point of the image is suppressed, the defect is detected with high accuracy. That is, the value of the difference signal becomes large due to a shift between a position separated by one cycle and the position of the sampling point (particularly a position shift at an edge portion of an image) and noise, and the influence of the position shift and the noise becomes a binary defect as a pseudo defect. Although this pseudo defect may appear in the binarized data, it usually appears in a different place, and is removed by taking a logical product of the binarized data.

[0004]

In the above-described conventional defect detection method, the logical product data is obtained from the image data of three consecutive repeating units, but the second type is capable of detecting the defect. Only the (center) repetition unit cannot detect defects present in the first and last repetition units of the repetition pattern. Also,
Japanese Patent Application Laid-Open No. 61-205811 does not disclose another specific means for detecting a defect existing in the first and last repeating units.

On the other hand, the applicant of the present application has disclosed in Japanese Patent Application No. 308208, which has been already filed, a delay circuit for delaying a signal by one cycle, and using a delay circuit for the first of three consecutive repeating units. By taking the logical product of the binary data obtained from the second repeating unit and the inverted data of the binary data obtained from the second and third repeating units, the first repeating unit Detect existing defects and obtain 2 from the second and third repeat units.
Periodic pattern for detecting a defect existing in the last repetition unit by taking the logical product of the binarized data and data obtained by inverting the value of the binarized data obtained from the first and second repetition units Discloses an inspection apparatus. However, this inspection apparatus is different from the defect detection method disclosed in Japanese Patent Application Laid-Open No. 61-205811 in that data obtained by inverting the value of the binarized data (inversion) in the detection of defects in the first and last repetition units. Since the subsequent data is logically ANDed using a part other than the true defect and the pseudo defect, the pseudo defect caused by the displacement of the sampling points and noise is not removed. It cannot be detected accurately.

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and an object of the present invention is to accurately detect defects at any point from the beginning to the end of a repetitive pattern of an inspection object. An object of the present invention is to provide a periodic pattern inspection device that can be detected by the above.

[0007]

A first periodic pattern inspection apparatus according to the present invention, which has been made to solve the above-mentioned problems, has a feature that a pattern in which a repetition unit is repeated by N (natural number) at a constant repetition period. A periodic pattern inspection apparatus for detecting a defect of an inspection object having: a) image input means for reading the pattern for each pixel and sequentially outputting a first image signal; and b) repeating the first image signal. Signal delay means for generating a second image signal delayed by one cycle of the cycle, and a third image signal obtained by delaying the first image signal by two cycles of the repetition cycle; c) the first image signal; A first difference signal that is a difference between the second image signal, a second difference signal that is a difference between the second image signal and the third image signal, and a third difference signal that is a difference between the third image signal and the first image signal.
A difference signal generating means for generating a difference signal; d) comparing the first difference signal with the second difference signal,
By generating a first detection signal indicating the presence or absence of a defect in each of the second to (N-1) th repeating units of the repeating pattern, and comparing the second difference signal with the third difference signal,
By generating a second detection signal indicating the presence / absence of a defect in each of the first to N-2th repeating units of the repeating pattern, and comparing the third difference signal with the first difference signal,
Defect detection means for generating a third detection signal indicating the presence or absence of a defect in each of the third to Nth repetition units of the repetition pattern; e) first to third detection signals for the first to third detection signals; Signal selection means for preventing overlapping of defect detection in the Nth repetition unit.

Further, the second periodic pattern inspection apparatus according to the present invention is the first periodic pattern inspection apparatus, wherein the first difference signal and the second difference signal, the second difference signal and the third difference signal, And a type determination unit for determining the type of the defect by comparing the two difference signals of the third difference signal and the first difference signal in consideration of the sign of each difference signal.

[0009]

In the first periodic pattern inspection apparatus according to the present invention, the image input means reads a repetitive pattern of the inspection object for each pixel and sequentially outputs a first image signal representing the pattern. From the first image signal, a second image signal obtained by delaying the first image signal by one repetition period and a third image signal obtained by delaying the second image signal by two periods are generated by a signal delay unit. The signal generating means generates a first difference signal from the first image signal and the second image signal, a second difference signal from the second image signal and the third image signal, and a third difference signal from the third image signal and the first image signal. Signals are respectively generated.
The values of these difference signals are small if no defect is present (ideally 0), but are large where a defect is present (peaks corresponding to the defect appear in the difference signal). Also,
Since the first image signal is used to generate the first difference signal and the third difference signal, if a defect appears in the image represented by the first image signal at a certain time, the peak corresponding to the defect is at the time Appears in both the first difference signal and the third difference signal. Similarly, when a defect appears in the image represented by the second image signal at a certain time, peaks corresponding to the defect appear at both times in the first difference signal and the second difference signal at a certain time. Third
If a defect appears in the image represented by the image signal, a peak corresponding to the defect is generated at that time by the second difference signal and the third signal.
Both appear in both of the difference signals.

The defect detection means generates first to third detection signals by comparing the first to third difference signals. That is, by comparing the first difference signal and the second difference signal, a first detection signal indicating the presence or absence of a defect is generated. The first detection signal obtained in this manner indicates whether or not a defect appears in the image represented by the second image signal. The generation of the first detection signal is one cycle longer than the second image signal in addition to the second image signal. A first image signal advanced by an amount and a third image signal delayed by one period from the second image signal are used. Therefore, the first detection signal indicates the presence / absence of a defect in each of the second to (N-1) th repetition units of the repetition pattern of the inspection object. Similarly to the first detection signal, the defect detection means compares the second detection signal by comparing the second difference signal with the third difference signal, and compares the third difference signal with the first difference signal. A third detection signal is generated. The second detection signal indicates whether a defect appears in the image represented by the third image signal, and the third detection signal indicates whether a defect appears in the image represented by the first image signal. Therefore, the second detection signal indicates the presence or absence of a defect in each of the first to N-2th repeating units, and the third detection signal indicates the presence or absence of a defect in each of the third to Nth repeating units. .

[0011] The signal selection means prevents duplicate detection of the defect, since there are duplicate detection signals with respect to the first to third detection signals obtained as described above. As described above, the first periodic pattern inspection apparatus detects a defect related to any repetition unit from the beginning to the end of the repetition pattern. Further, when detecting a defect with respect to the repeating units at both ends, it is not necessary to use inverted data of the binarized data.

In the second periodic pattern inspection apparatus according to the present invention, the type judging means performs comparison between the first to third difference signals by taking into account the sign of the difference signals. Determine the type of defect. That is, if the first difference signal is generated by subtracting the second image signal from the first image signal, the value of the first difference signal is determined by the white defect (the signal value is large and the image is regarded as a bright portion) appearing in the second image signal. The portion corresponding to the defect which appears in the image is negative, and the portion corresponding to the black defect (a defect which has a small signal value and appears in the image as a dark portion) is positive. On the other hand, when the second difference signal is generated by subtracting the third image signal from the second image signal, the portion corresponding to the white defect appearing in the second image signal is positive, and the value of the second difference signal is black. The portion corresponding to the defect is negative. Therefore, by comparing the first difference signal and the second difference signal, it is determined that a defect existing in a portion where the value of the first difference signal is negative and the value of the second difference signal is positive is a white defect. The defect existing in the portion where the value of the first difference signal is positive and the value of the second difference signal is negative is determined to be a black defect. Similarly, by comparing the second difference signal and the third difference signal, it is determined that a defect existing in a portion where the value of the second difference signal is negative and the value of the third difference signal is positive is a white defect. Then, it is determined that a defect existing in a portion where the value of the second difference signal is positive and the value of the third difference signal is negative is a black defect (however, the first image signal is subtracted from the third image signal). Third
A difference signal is generated). Further, the third difference signal is compared with the first difference signal, and it is determined that a defect existing in a portion where the value of the third difference signal is negative and the value of the first difference signal is positive is a white defect. The defect existing in a portion where the value of the third difference signal is positive and the value of the first difference signal is negative is determined to be a black defect.

[0013]

FIG. 1 is a block diagram showing an entire configuration of a periodic pattern inspection apparatus according to one embodiment of the present invention. The inspection apparatus includes an image input unit 50, first and second period delay units 11 and 12, first, second and third comparison units 21 and 2,
2 and 23, a defect determination unit 30, a microprocessor (MPU) 70, a CRT 71 and a keyboard 72 connected thereto.

In the image input unit 50, the stage 51 is moved in the sub-scanning direction Y by the stage drive system 54 controlled by the MPU 70 via the control system 55. On the stage 51, an inspection object 61 having a pattern repeated at a constant repetition cycle is placed, and an image of this repetition pattern is transferred to the stage 5 in the sub scanning direction Y.
In the moving process 1, the reading device 52 reads the image data in the main scanning direction X in pixel units. The signal representing the read image is converted into, for example, an 8-bit digital signal by the A / D converter 53 and output from the image input unit 50. The first period delay unit 11, the first comparison unit 21, and the second
The data is input to the comparison unit 22.

The first cycle delay unit 11 delays the image signal (hereinafter, referred to as “current signal”) SR output from the image input unit 50 by one cycle of the repetition pattern of the inspection object 61 (hereinafter, referred to as “signal”). SQ
Generate The one-cycle delay signal SQ includes the second and third signals.
An image signal (hereinafter, referred to as a “two-cycle delay signal”) SP that is input to the comparison units 22 and 23 and input to the second cycle delay unit 12 and is further delayed by one cycle is generated. The two-period delay signal SP is supplied to the third and first comparing units 2
3 and 21 are input.

The first comparator 21 subtracts the two-cycle delay signal SP from the current signal SR to obtain a difference signal SN = SR-SP, and the second comparator 22 subtracts the current signal SR from the one-cycle delay signal SQ to obtain a difference signal. ST = SQ−SR, and the third comparing unit 23 subtracts the one-cycle delay signal SQ from the two-cycle delay signal SP to obtain a difference signal S
S = SP−SQ is generated respectively. These difference signals S
N, ST, and SS are input to the defect determination unit 30.

The values of the difference signals SN, ST, SS are ideally zero if there is no defect due to the periodicity of the pattern of the object to be inspected. Therefore, the presence or absence of a defect can be detected. However, as in the conventional case (see FIG. 7D),
The peak corresponding to one defect has one cycle (difference signal SS, ST
) Or two periods (in the case of the difference signal SN) appearing two times apart, it is necessary to identify which of the repetition units that is one or two periods apart has a defect. The defect determining unit 30 detects the presence or absence of a defect and identifies the position of the defect by performing the above-described identification.

FIG. 2 is a schematic diagram showing the principle of defect detection (including position identification) in the defect determination section 30. FIG.
(A) shows an image of three continuous repeating units in the repeated pattern on the inspection object 61. Therefore,
These images P, Q, and R correspond to a certain one of the repeating patterns.
It is an image represented by a two-cycle delay signal SP, a one-cycle delay signal SQ, and a current signal SR during a period of a cycle. As shown in FIG. 2A, the image P has black defects (hatched portions), the image Q has both white defects (white portions) and black defects, and the image R has white defects. Each exists. In these images, a satin pattern portion surrounded by a rectangle represents an image region having an intermediate gradation value. FIG. 2B illustrates a difference image S obtained by subtracting the image Q from the image P, a difference image T obtained by subtracting the image R from the image Q, and a difference image N obtained by subtracting the image P from the image R. Then, the absolute value of each pixel of the difference images N, S, and T is compared with a predetermined threshold.
An image U obtained by binarizing the difference images N, S, and T so that the difference image N, S, and T become “0” at a location equal to or less than a predetermined threshold.
V and W are as shown in FIG. Defects appearing in image P appear in both binary images U and V, defects appearing in image Q appear in both binary images V and W, and defects appearing in image R appear in both binary images W and U, respectively. .
Therefore, by taking the logical product of the binary images U and V, only the defects present in the image P appear, and by taking the logical product of the binary images V and W, only the defects present in the image Q are found. By taking the logical product of the binary images W and U that appear, binary images in which only defects present in the image R appear can be obtained.

In this embodiment, in order to detect a defect based on the above principle, the defect judging section 30 has a configuration shown in FIG. That is, the defect determination unit 30 includes the first, second, and third comparators 31, 32, and 33, and the AND gates 35, 3
6 and 37, and the difference signals SS, ST, and SN are input to first, second, and third comparators 31, 32, and 33, respectively. Each of the comparators 31, 32, and 33 also receives an inspection condition setting signal Sset indicating a predetermined threshold. As the inspection condition setting signal Sset, for example, a configuration may be used in which a threshold value can be set in a predetermined register from the MPU 70, and an output signal of the register may be used. Each comparator 31, 32, 33
Compares the absolute value of each of the difference signals SS, ST, and SN with a threshold value indicated by the inspection condition setting signal Sset. If the absolute value of the difference signal is greater than the threshold value, the value becomes “1”;
If the value is equal to or smaller than the threshold value, signals SV, SW, and SU that become "0" are output. Among the binarized signals (binary signals), the binary signal SV output from the first comparator 31 is A
The binary signal SW output from the second comparator 32 is supplied to the ND gates 35 and 36 and supplied to the AND gates 36 and 37 by the third signal.
The binary signal SU output from the comparator 33 is input to AND gates 37 and 35, respectively. Each of these AN
Control signals C1, C2, and C3 described below are also input to the D gates 35, 36, and 37, respectively.

As shown in FIG. 4A, the inspection object 61
Is n, the repetition period is L, and the direction from left to right in the drawing is the scanning direction. In the present embodiment, given the region of the repetitive pattern and the repetition period L in the inspection object 61, the MPU 70 determines the timing of the control signals C1, C2, and C3 as shown in FIGS. (Timing at which the signal value changes) is calculated, and data indicating these timings is set in a register (not shown) in the control system 55. That is, assuming that a period of one cycle in which the current signal SR represents the image of the i-th repetition unit is called an "i-th period", the control signal C1 is a third period, and the control signal C2 is a third to a third period. During the n-th period, the control signal C3 becomes'
It becomes 1 and all control signals become 0 in other periods. Therefore, the MPU 70 sets data indicating the start and end timings of the third period and the n-th period in the register. The control system 55 generates control signals C1, C2, and C3 based on the setting data and inputs the generated signals to the defect determination unit 30.

Attention is now directed to a period of one cycle in which the current signal SR represents an image of the third repetition unit (hereinafter referred to as a “current period”).
Represents the image of the second repetition unit, and the two-cycle delayed signal S
P represents the image of the first repeating unit. Therefore, in the current period, the images of the first, second, and third repetition units can be associated with the images P, Q, and R in FIG. When associated in this way, the difference signals S S, S T, S N correspond to the difference images S, S in FIG.
T, N, respectively, and the binary signals SV, SW, SU are:
These correspond to the binary images V, W, and U in FIG. Therefore, the logical product of the binary signals SU and SV corresponds to the binary image PD in which only the defect appearing in the image P appears (see FIG. 2D), and indicates the defect existing in the first repeating unit. become. Similarly, the logical product of SV and SW and the logical product of SW and SU correspond to the binary images QD and RD in which only the defects appearing in the images Q and R respectively correspond to the second and third images.
It will indicate a defect present in the th repeat unit. The same association can be made in a period (fourth period) for the next cycle of the current period. The logical product of SU and SV, the logical product of SV and SW, and the logical product of SW and SU are Defects present in the second, third and fourth repeat units are shown. Considering the same hereinafter, in the i-th period, the logical product of SU and SV, S
The logical product of V and SW and the logical product of SW and SU are i−
It can be seen that it indicates a defect existing in the second, i-1st, and ith repeating units (where i = 3, 4,..., N).

As shown in FIG. 3, the AND gate 35 outputs the signal SPD of the logical product of the signals SU, SV and C1, and the AND gate 36 outputs the signal SQD of the logical product of the signals SV, SW and C2.
The ND gate 37 outputs a signal SR of the logical product of the signals SW, SU and C3.
D is output. Therefore, the logical product between the three binary signals SV, SW, SU and the control signals C1, C2, C3
As can be seen from the above description of the signal SPD
The presence or absence of a defect in the first repeating unit during the period is determined by a signal S
QD indicates the presence or absence of a defect in the second to (n-1) th repeating units in the third to nth periods, and the signal SRD indicates the presence or absence of a defect in the nth repeating unit in the nth period (FIG. 4E).
), And all of these signals become '0' in other periods. By using such signals SPD, SQD, and SRD, it is possible to detect a defect in the inspection object 61 and to specify which part of the repetition unit from the beginning to the end of the repetition pattern has a defect. As can be seen from the above, the control signals C1, C2, and C3 select a period in which each of the logical product of SU and SV, the logical product of SV and SW, and the logical product of SW and SU is used as a defect detection signal, Defects in each repetition unit from the beginning to the end are detected without duplication.

According to the present embodiment as described above, the signal SP
D and SRD make it possible to identify a defect in the first (first) repeating unit and a defect in the last (n-th) repeating unit of the repeating pattern of the inspection object 61.
Also, in the generation of the signals SPD and SRD for detecting the defect in the first and last repetition units, similarly to the generation of the signal SQD for detecting the defect in the other repetition units, the signal obtained by binarizing the absolute value of the difference signal is used. Since the logical AND is used, the influence of the displacement and noise of the sampling points of the image can be suppressed.
That is, the binarized signals SV, SW, and SU are equal to the difference signal S.
If the absolute value of S, ST, SN is larger than a predetermined threshold,
1), and if it is equal to or less than a predetermined threshold value, it becomes '0' (because it becomes '1' only in the portion corresponding to the defect), and the influence of the displacement of the sampling point of the image or the influence of noise is regarded as a pseudo defect of 2
Even if it appears in the value signals SV, SW, and SU (the location where this pseudo defect appears usually depends on the binary signal), the pseudo defect is removed by taking a logical product.

As described above, the presence / absence of a defect can be detected together with the position of the defect by the defect determination unit 30 in the inspection apparatus of the present embodiment. However, in the circuit shown in FIG. A black defect cannot be identified. However, it is possible to specify the defect type based on a principle similar to the principle shown in FIG. That is, as shown in FIG. 2, the defect appearing in the image P appears in both the difference images N and S, and the value of the image portion corresponding to the defect is negative in the difference image N if the defect is a white defect. In the case of the image S, it is positive, and in the case of a black defect, it is positive in the difference image N and negative in the difference image S. Therefore, the common part of the negative part in the difference image N and the positive part in the difference image S represents a white defect appearing in the image P (in the example of FIG. 2, since no white defect appears in the image P, this common part is Absent), and the common part of the positive part in the difference image N and the negative part in the difference image S represents a black defect appearing in the image P. Therefore, in order to identify the type of the defect appearing in the image P, the white defect and the black defect may be separately detected by obtaining these common parts. Similarly, the defect appearing in the image Q appears in both the difference images S and T, and the value of the image portion corresponding to the defect becomes negative in the difference image S and positive in the difference image T if it is a white defect, If the defect is a black defect, the difference image S is positive and the difference image T is negative. Therefore, in order to identify the type of the defect appearing in the image Q, a common part between the negative part in the difference image S and the positive part in the difference image T is obtained, and only the white defect is detected, and the positive part in the difference image S is detected. Then, only the black defect may be detected by finding the common portion between the portion and the negative portion in the difference image T. Further, the defect appearing in the image R appears in both the difference images T and N, and the value of the image portion corresponding to the defect is negative in the difference image T, positive in the difference image N, and black in the case of a white defect. In the case of a defect, the difference image T is positive and the difference image N is negative. Therefore, the image R
In order to identify the type of defect appearing in the difference image T, a common part between the negative part in the difference image T and the positive part in the difference image N is determined, and only the white defect is detected. What is necessary is just to find the common part with the negative part in the image N and detect only the black defect.

FIG. 5 shows an example of a defect type detection circuit for specifying a defect type based on the above principle. The signals Ssgn, Nsgn, and Tsgn input to this circuit are sign signals indicating the sign of the difference signals SS, ST, and SN, respectively. If the value of the difference signal is negative, the sign signal is “1”, and the difference signal is “1”. Is positive, the sign signal is '0'. Here, the sign signal Ssgn
Is a relationship including the negative portion of the difference image S in FIG. 2B (ideally, these image portions match, but in reality, the difference Since the value of the signal SS does not become 0 even in a portion other than the defect and is either positive or negative, the image portion where Ssgn is “1” includes the negative portion of the difference image S). For this reason, according to the above principle, the white defect among the defects appearing in the image P of FIG. 2A is included in the image portion where the signal PW, which is the logical product of the inverted signal of Ssgn and Nsgn, is “1”. The black defect is included in the image portion where the signal PB, which is the logical product of the inverted signal of Nsgn and Ssgn, is "1". Similarly, among the defects appearing in the image Q, the white defect is Tsgn
The signal QW, which is the logical product of the inverted signal of Ssgn and Ssgn, is included in the image portion where the value is “1”.
The signal QB, which is the logical product of gn and gn, is included in the image portion where it is "1". Further, among the defects appearing in the image R, white defects are as follows.
The signal RW, which is the logical product of the inverted signal of Nsgn and Tsgn, is
The black defect is included in the image portion where the signal RB, which is the logical product of the inverted signal of Tsgn and Nsgn, is '1'. Therefore, when a defect is detected by the signals SPD, SQD, SRD output from the circuit shown in FIG. 3, the signals PW, PB, QW,
By examining the values of QB, RW, and RB, it is possible to identify whether the defect is a white defect or a black defect. Therefore, when it is necessary to specify not only the position of the defect but also the type of the defect, such a defect type detection circuit may be incorporated in the defect determination unit 30.

As described above, the signals SPD, SQD and SRD indicating the presence / absence of a defect and its position are output from the defect judging section 30 (if the defect type detection circuit shown in FIG. Signals PW and P indicating the presence or absence of a defect for each type
B, QW, QB, RW, and RB are also output), and these signal values are taken into the MPU 70 as defect data. MP
U70 displays a defect image and coordinates on the CRT 71 based on the defect data (signals PW, PB, QW, QB, R
When the values of W and RB are also taken into the MPU 70,
The defect type is also displayed). The MPU 70 operates based on an input operation with the keyboard 72, displays information on such a defect, and operates the stage 5 as described above.
1 is controlled.

[0027]

According to the present invention, it is possible to detect whether or not a defect exists in the first and last repetition units of a repetition pattern of an inspection object. Also, when detecting defects for the first and last repetition units, the defect is detected based on the comparison of the difference signals without using inverted data of the binary data. The effect of noise is suppressed. Therefore, in the inspection of the inspection object having the periodic pattern, it is possible to detect with high accuracy any defect from the beginning to the end of the repetitive pattern.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a periodic pattern inspection apparatus according to one embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating the principle of defect detection.

FIG. 3 is a circuit diagram showing a configuration of a defect determination unit in the periodic pattern inspection apparatus according to one embodiment of the present invention.

FIG. 4 is a diagram showing a control signal for selecting a signal for detecting a defect.

FIG. 5 is a diagram showing a configuration of a defect type detection circuit for specifying a defect type.

FIG. 6 is a block diagram showing a configuration of a conventional periodic pattern inspection device.

FIG. 7 is a diagram for explaining the principle of a conventional periodic pattern inspection device.

FIG. 8 is a diagram for explaining the principle of a conventional defect detection method.

[Explanation of symbols]

 Reference Signs List 11 first period delay unit 12 second period delay unit 21 first comparison unit 22 second comparison unit 23 third comparison unit 30 defect determination unit 50 image input unit 70 MPU SR, SQ, SP ... Image signal SS, ST, SN ... Difference signal C1-C2 ... Control signal

──────────────────────────────────────────────────続 き Continuing on the front page (72) Kunio Ueda 4-chome Tenjin, Horikawa-dori-Terauchi, Kamigyo-ku, Kyoto 1 Daikita Screen Manufacturing Co., Ltd. (72) Inventor Yutaka Konaka Kyoto 4-chome Tenjin, Horikawa-dori-Teranouchi, Kamigyo-ku, Tokyo 1 Daikoku Screen Manufacturing Co., Ltd. (56) References JP-A-7-270334 (JP, A) JP-A-7-159342 (JP) JP-A-7-270333 (JP, A) JP-A-6-74099 (JP, A) JP-A-4-316346 (JP, A) JP-A-59-192943 (JP, A) 6-129992 (JP, A) JP-A-61-205811 (JP, A) JP-A-3-30878 (JP, A) JP-A-62-245484 (JP, A) JP-A-6-18443 (JP, A) A) Japanese Patent Application Laid-Open No. 6-74995 (JP, A) (58) nt.Cl. ⁶ , DB name) G01N 21/88-21/89

Claims (2)

(57) [Claims] 1. A periodic pattern inspection apparatus for detecting a defect of an inspection object having a pattern in which a repetition unit is repeated by N (natural number) at a constant repetition cycle, a) reading said pattern for each pixel and Image input means for sequentially outputting image signals; b) a second image signal obtained by delaying the first image signal by one cycle of the repetition cycle; and an image input means for delaying the first image signal by two cycles of the repetition cycle. C) a first difference signal which is a difference between the first image signal and the second image signal, and a second signal which is a difference between the second image signal and the third image signal. A third signal which is a difference signal and a difference between the third image signal and the first image signal.
A difference signal generating means for generating a difference signal; d) comparing the first difference signal with the second difference signal,
By generating a first detection signal indicating the presence or absence of a defect in each of the second to (N-1) th repeating units of the repeating pattern, and comparing the second difference signal with the third difference signal,
By generating a second detection signal indicating the presence / absence of a defect in each of the first to N-2th repeating units of the repeating pattern, and comparing the third difference signal with the first difference signal,
Defect detection means for generating a third detection signal indicating the presence or absence of a defect in each of the third to Nth repetition units of the repetition pattern; e) first to third detection signals for the first to third detection signals; A signal selection unit for preventing overlapping of defect detection in the Nth repetition unit. 2. The two difference signals, a first difference signal and a second difference signal, a second difference signal and a third difference signal, and a third difference signal and a first difference signal, are used to determine whether the value of each difference signal is positive or negative. 2. The periodic pattern inspection apparatus according to claim 1, further comprising: a type determination unit configured to determine a type of the defect by performing a comparison while taking into account.