JPH01255073A - Method and device for checking pattern - Google Patents

Abstract

PURPOSE:To detect the parts of pictures as defects by detecting a higher density part in the high density part, and a lower density part in the low density part as the pictures. CONSTITUTION:A control device 5 reads a television picture at a standard pattern stored into a picture memory 31. At such a time, the binarization picture at the standard pattern obtained by binarizing the part corresponding to a check object pattern DP in the visual field of a television camera 2 by means of a third slicer 43 is stored into the memory 31. Thereafter, the pattern DP is image-picked-up by the camera 2. A picture signal obtained by the image pickup is once store din the memory 31, successively converted into the binarization picture by respective slicers 41-43, and stored into the memory 31 again. At such a time, by properly selecting the slice level of the respective slicers 41-43, the higher concentration part in the high density part and the lower density part in the low density part can be detected as the defects.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method and apparatus for detecting pattern defects, and more specifically, a method for inspecting a pattern printed on an IC board, etc., and a method used therefor. Regarding equipment.

[Prior art] An IC (integrated circuit) substrate is produced by applying a resist onto a silicon substrate, exposing it to light after closely adhering it to a photomask on which a wiring pattern, etc. has been formed in advance, and then developing it. However, as ICs have become more dense and multi-layered in recent years, defect inspection of the photomask patterns mentioned above or wafer patterns formed on silicon substrates has become increasingly important for product quality control and productivity control. This has become an important issue from this point of view.

By the way, inspection for pattern defects as described above has conventionally been carried out by human visual inspection in boat fishing.

However, with this method, it cannot be said that no pattern defects are overlooked, resulting in a decrease in the yield and reliability of finished products, and furthermore, it cannot be said that the inspection efficiency and productivity are high.

Under these circumstances, the applicant of the present application has previously proposed the invention "Pattern inspection method and apparatus" disclosed in Japanese Patent Application Laid-Open No. 60-200382, which is an invention of the inventors of the present application.

In short, this invention detects defects in the pattern to be inspected by using a defect-free pattern as a reference pattern and calculating the difference between a binary image of this pattern and a binary image of the pattern to be inspected. .

[Problem to be solved by the invention]

By the way, in the above-mentioned invention, since the binarization of the image is performed using a threshold value that represents the density between the pattern part and the background substrate part, the disulfide threshold value that exists in the part where the density is higher than the binarization threshold value is Defects with low concentration and vice versa are easily detectable. However, on the photomask pattern or wafer pattern, if there are defects with a lower concentration in the part where the density is lower than the binarization threshold, or if there are defects with a higher density in the part with the density higher than the binarization threshold. However, such defects cannot be detected by the above-described invention.

Specifically, it is as shown in FIG. That is, the pattern to be inspected DP includes a wiring pattern 52 with a slightly lower concentration formed on a substrate 51 with a slightly higher concentration, and whether or not this wiring pattern 52 is formed accurately is the object of inspection. An image obtained by binarizing this pattern DP to be inspected using a density intermediate between the density of the substrate 51 and the density of the wiring pattern 52 as a threshold value is as shown by DPO in FIG.

On the other hand, an image obtained by binarizing the standard pattern sp, which has no defects, in the same manner as the pattern to be inspected DP, becomes SPo in FIG. The image obtained by calculating the logical sum of each pixel of the binarized image spo of this standard pattern SP is shown in FIG. 9 SP7! The enlarged image as shown in Figure 9 and the logical product image are shown in SP.
This results in a reduced image as shown in s.

Then, by subtracting the enlarged binary image spz of the standard pattern sp from the binary image DPO of the pattern to be inspected DP, a surplus image as shown by Wl in FIG. 9 is obtained. represents a short circuit or protrusion, etc. In addition, the reduced binarized image SP of the standard pattern SP
When the binarized image DPO of the pattern DP to be inspected is subtracted from s, a missing image as shown by -2 in FIG. 9 is obtained.
This represents a missing defect, such as a disconnection or defect in the wiring.

In such a conventional method, defects 53 and wiring patterns 52 that are present on the substrate 51 and have a higher concentration than the substrate 51 are removed.
Defects 54 having a lower concentration than the wiring pattern 52 existing above are detected on the substrate 51 when the pattern to be inspected DP is binarized.
Alternatively, since they are binarized to the same value as the wiring pattern 52, they become indistinguishable. Therefore, with such a conventional method, it is impossible to detect the defects 53 which are present on the substrate 51 and have a higher concentration than the substrate 51, and the defects 54 which are present on the wiring pattern 52 and have a lower concentration than the wiring pattern 52.

The present invention detects defects in photomask patterns or evaporation patterns that exist in areas with high concentrations that cannot be detected by comparing the inspection target pattern with a standard pattern by simple binarization. It is an object of the present invention to provide a pattern inspection method and apparatus that can easily and accurately detect defects with a high density or defects with a lower density existing in a lower density part.

[Means to solve the problem]

Conventionally, the pattern inspection method and apparatus of the present invention simply
The image of the pattern to be inspected, which has been converted into a value, can be obtained as an image of a part where the density is higher (or lower) than the substrate part and a part where the density is lower (or higher) than the pattern part. .

The present invention provides a pattern inspection method for detecting defects in a pattern to be inspected in which a pattern having a lower (or higher) concentration than that of the substrate is formed on a substrate. A portion detected by an image binarized with a threshold value representing 4 degrees (or higher) and a portion obtained by an image binarized with a threshold value representing a higher (or lower) density than the substrate portion of the pattern to be inspected. It is characterized by detecting as a defect.

[Effect]

In the pattern inspection method and apparatus of the present invention, since a higher density part of a higher density part and a lower density part of a lower density part are obtained as images, these parts can be detected as defects. .

[Principle of the invention]

For example, as shown in FIG. 2, a relatively high concentration substrate 5
A wiring pattern 52 with relatively low concentration is formed on 1,
There are defects 53 on the substrate 51 that have a higher concentration than the substrate 51.
Further, when there is a pattern to be inspected DP in which a defect 54 having a lower concentration than the wiring pattern 52 exists on the wiring pattern 52 and a standard pattern SP as shown in FIG. As shown in the figure (
become.

That is, the concentration distribution of the inspection target pattern DP as shown in FIG. 2 has a peak P54 corresponding to the defect 54 with the lowest concentration.
, a peak P52 corresponding to the wiring pattern 52 with a relatively low concentration, and a peak P corresponding to the substrate 51 with a relatively high concentration.
51, peak P53 corresponding to defect 53 with the highest concentration
4 peaks are shown. On the other hand, the concentration distribution of the standard pattern sp as shown in FIG. 3 shows two peaks: a peak P52 corresponding to the wiring pattern 52 with a relatively low concentration, and a peak P51 corresponding to the substrate 51 with a relatively high concentration.

Conventionally, the value between the peak P52 corresponding to the wiring pattern 52 and the peak P51 corresponding to the board 51 is set as a threshold value ↑■0
Since the peaks P54 and P53 corresponding to defects 54 and 53 as shown in FIG.
were detected as having the same value as the wiring pattern 52 or the substrate 51, respectively, and could not be detected as a defect.

However, in FIG. 4, the peak P corresponding to the defect 54
54 and the peak P52 corresponding to the wiring pattern 52 is obtained as the first threshold value THI to obtain a binary image, it is possible to distinguish between the wiring pattern 52 and the defect 54 whose density is lower than this, Also, peak P53 corresponding to defect 53
If a binary image is obtained by setting a value between the peak P51 and the peak P51 corresponding to the substrate 51 as the second threshold TH2, it becomes possible to discriminate between the substrate 51 and the defect 53 whose density is higher than this.

The above is the principle of the present invention, but the relationship between the concentrations of the substrate 51 and the wiring pattern 52 is opposite to the above explanation, that is, the wiring pattern 5 has a relatively high concentration on the substrate 51 with a relatively low concentration.
2 is formed, it is possible in principle to detect each defect in the same way.

[Embodiments of the invention]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on drawings showing embodiments thereof. However, in the embodiment described in detail below, it is assumed that the substrate 51 of the pattern to be inspected DP has a relatively higher concentration than the wiring pattern 52.

FIG. 6 is a block diagram showing the configuration of an apparatus used to implement the pattern inspection method according to the present invention.

Reference numeral 1 in the figure is an inspection table, which includes a fixed base IB horizontally fixed on a base (not shown) and an X stage IX supported on the fixed base IB and moved in the depth direction (X direction) in the figure. , Y stage IY is supported by this X stage IX and moves in the left-right direction (Y direction) perpendicular to the X direction in the figure, and furthermore, it is fixed to fixed base IB and X stage IX is moved to fixed base IB. The Y stage drive motor MX is configured to move the Y stage IY in the Y direction, and the Y stage drive motor MY is fixed to the X stage IX and is fixed to the A pattern to be inspected DI', which is an object to be inspected, is placed thereon.

Therefore, the pattern to be inspected DP placed on the Y stage IY can be moved in two-dimensional directions.

Note that the X stage drive motor MX and Y stage drive motor MY are driven by respective drive circuits DX and DY.

A two-dimensional imaging device, such as a television camera 2, is fixed above the examination table 1 with the optical axis of its lens system pointing vertically downward via an appropriate vibration isolator (not shown).

This television camera 2 has autofocus (AP)
) A sensor 2S and an AP drive motor 2M are provided, and when the detection signal of the AF sensor 2S is given to the needle control device 2C, a control signal is outputted from the needle control device 2C to the drive circuit 2D, and thereby the AP drive The motor 2M is driven to perform autofocus control.

An image signal captured by the television camera 2 is given to an image processing device 3.

The image processing device 3 includes an image memory 31 and an enlargement/reduction circuit 3.
2, an image calculation circuit 33, an image processing controller 34, slicers 41 to 43, and the like.

This image processing device 3 stores an image of the pattern to be inspected DP captured by the television camera 2 in an image memory 31.
This image is once stored in the first to third slicers 41 to 43, each having a different threshold value as the slice level.
The images are digitized and each binarized image is stored in the image memory 31. At this time, the slice levels of the first to third slicers 41 to 43 are the first threshold value THI, the second threshold value TH2, and the reference threshold value THO, which have already been described in the [Principle of the Invention] section.
It is.

The slice level of the first slicer 41 is set to the aforementioned first threshold value Tl+1, that is, the value representing the concentration between the wiring pattern 52 and the defect 54 whose concentration is lower than this. The binarized image DPI represents a wiring pattern 52 and a defect 54 with a lower density than that existing in this wiring pattern 52. Further, the slice level of the second slicer 42 is set to the aforementioned second threshold TI (2, that is, a value representing the concentration between the substrate 51 and the defect 53 having a higher concentration than this, and this second A binarized image DP2 obtained by slicing +42 represents the substrate 51 and a defect 53 with a higher density present in the substrate 51.Furthermore, the slice level of the third slicer 43 is set to the reference threshold value THO described above.
The binarized image DPO produced by this third slicer 43 represents the substrate 51 and the wiring pattern 52.

On the other hand, the standard pattern SP is also stored in the image memory 31 in advance, and the image processing device 3 converts the standard pattern SP into a binary image SPO by the third slicer 43, and further converts the standard pattern SP into a standard image by the enlarging/reducing circuit 32. The pattern sp is enlarged and reduced, and an enlarged binarized image SP7 of this standard pattern sp!
Defects in the pattern to be inspected DP are detected by calculating the reduced binarized image SPs and the binary image DPO based on the reference threshold value THO of the pattern to be inspected DP in the image calculation circuit 33.

In addition, the control device 5 gives a command to the image calculation circuit 33 to correct the position of the pattern to be inspected DP within the field of view of the television camera 2 prior to the above-mentioned image processing, and also transmits the above-mentioned X stage drive motor M
Drive circuits DX and D for X and Y stage drive motors MY
By driving and controlling Y, the pattern to be inspected DP is sequentially moved in both the x and y directions while
The pattern to be inspected DP is imaged. Further, the control device 5 gives a command to the above-mentioned image processing device 3 to process the image captured by the television camera 2.

Note that the control device 5 stores in advance data for various controls as described above in a built-in storage device such as a ROM.

Further, the image display device 7 appropriately displays various images stored in the image memory 31, and the display device 8 performs other various displays.

Next, the operation of the apparatus of the present invention configured as described above will be explained below with reference to the flowchart of FIG. 7 showing the inspection procedure and the schematic diagrams of various patterns shown in FIG.

First, the control device 5 reads the traverse data of the inspection table 1, and moves the inspection table 1 to the initial position based on this.

Next, the control device 5 reads out the television image of the standard pattern SP stored in the image memory 31 in advance, and selects the portion corresponding to the range of the pattern to be inspected DP that is within the field of view of the television camera 2 at that time. slicer 431
Binarized image spo of the standard pattern sp
is stored in the image memory 31.

Thereafter, the control device 5 reads various control data from the ROM and causes the television camera 2 to take an image of the pattern leaf to be inspected. The image signal obtained by this imaging is temporarily stored in the image memory 31, and is sequentially transferred to each slicer 41 to 43.
are converted into binary images and stored in the image memory 31.
will be restored.

By the way, it is now assumed that the pattern DP to be inspected is in a state as shown in FIG. That is, this pattern to be inspected DP
In this case, a bar-shaped wiring pattern 52 is located from the right end to the center of the substrate 51, and a defect 54 with a lower concentration than the wiring pattern 52 exists on the wiring pattern 52, and the substrate does not overlap with the wiring pattern 52. Defects 53 having a higher concentration than the substrate 51 exist on the substrate 51 . Furthermore, the wiring pattern 52 has a missing defect 55 on its upper side and a redundant defect 56 on its lower side.

In such a pattern to be inspected DP, a straight line A-A passes only through the substrate 51 and the defect 53, which has a higher concentration than the substrate 51.
and a wiring pattern 52 and a defect 5 with a lower concentration than this.
The concentration distribution along the straight line aa passing through 4 is as shown in FIG. That is, if the concentration of the substrate 51 shown by B-B is used as a reference, the concentration of defects 53 on straight line A-A is equal to the standard straight line B.
-Protrudes from B to the side with higher concentration, and the straight line a-a
The defect 54 in FIG.
It is even more prominent than b.

Therefore, the third slicer 43 sets the intermediate value between the density B-B of the substrate 51 and the density b-b of the wiring pattern 52 as the reference threshold value Tll0, and sets this reference threshold value THO as the slice level.
An image DP obtained by binarizing the image of the pattern DP to be inspected at
If O is obtained and the difference is taken between the reduced binary image SPs and the enlarged binary image 5PI2 of the standard pattern sp as in the prior art described above, the missing defect 55 and the surplus of the wiring pattern 52 on the inspection target pattern inspection are obtained. Defect 56 can be detected as a defect.

On the other hand, the intermediate concentration between the concentration peak bb of the wiring pattern 52 and the concentration peak C-C of the defect 54 having a lower concentration is set as a first threshold value THI, and the first threshold value T1 is set as the slice level. If an image DPI obtained by converting the pattern to be inspected DP into a binary value is obtained by the slicer 41 of No. 1, a defect 54 having a lower density on the wiring pattern 52 is detected.

Further, the intermediate concentration between the peak B-B of the substrate 51 and the peak I)-D of the defect 53, which has a higher concentration, is set as a second threshold T.
)+2, and if an image plane 2 is obtained by binarizing the inspection target pattern DP by the second slicer 42 with this second threshold value TH2 as the slice level, then the defect 53 on the substrate 51 with a higher density is obtained. is detected.

The actual procedure is to obtain binarized images by the first slicer 41 and the second slicer 42, and to remove both defects 53.
．． 54, and then the third slicer 53 binarizes the pattern DP to be inspected using the reference threshold THO ii! i
i-image DPI is obtained, and after the position is corrected by the image calculation circuit 33, the image town and -2 are obtained by calculating the difference between the enlarged binarized image spρ and the reduced binarized image SPs of the standard pattern sp, Perform defect judgment.

Thereafter, the control device 5 moves the X stage IX and the Y stage IY according to the traverse data, causes the television camera 2 to image the next field of view, and repeats the same image processing and result determination as described above.

In the above embodiment, an example in which the concentration of the substrate 51 is relatively higher than that of the wiring pattern 52 has been described, but the present invention is also applicable to a case where the concentration relationship is completely reversed.

Furthermore, in the above embodiment, a binarized image of the pattern to be inspected DP is obtained using three different threshold values;
Of course, a quaternary image may be obtained using three threshold values at the same time.

〔Effect of the invention〕

As described above, according to the present invention, not only missing portions and surplus portions of wiring patterns can be detected as defects, but also defects and wiring with higher (or lower) concentration on the substrate that could not be detected conventionally. It becomes possible to detect defects with a lower (or higher) concentration on a pattern, and more rigorous pattern inspection becomes possible.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of a pattern to be inspected as an inspection target according to the present invention and a density graph showing three threshold values for its binarization, FIG. 2 is a schematic diagram of the pattern to be inspected, and FIG. Figure 2 is a schematic diagram of a standard pattern corresponding to the pattern to be inspected, Figure 4 is a graph showing the density distribution of the pattern to be inspected in Figure 2, and Figure 5 is a graph showing the density distribution of the standard pattern in Figure 3. , FIG. 6 is a block diagram showing the configuration of the device of the present invention, FIG. 7 is a flowchart showing its operating procedure, FIG. 8 is a schematic diagram of various images similarly showing its operating procedure, and FIG. 9 is a diagram of the prior art. It is an explanatory diagram. 2... Television camera 3... Image processing devices 41, 42... Slicer 51... Board 52
...Wiring pattern 53...Defects with higher concentration 54...Defects with lower concentration

Claims (1)

[Claims] 1. In a pattern inspection method for detecting defects in a pattern to be inspected in which a pattern with a lower (or higher) concentration than that of the substrate is formed on a substrate, the pattern to be inspected is inspected from a pattern portion. A portion detected by an image binarized with a threshold value representing a low (or high) density and an image obtained by binarizing the inspection target pattern with a threshold value representing a higher (or lower) density than the substrate portion. A pattern inspection method characterized by detecting a portion as a defect. 2. An imaging device that images a pattern to be inspected in which a pattern with a density lower (or higher) than that of the substrate is formed on a substrate; and an image of the pattern to be inspected captured by the imaging device, a first image binarization circuit that binarizes an image of the pattern to be inspected captured by the imaging device using a first threshold representing a density lower (or higher) than the substrate part; a second image binarization circuit that binarizes the image using a second threshold representing (or low) density; A pattern inspection device characterized in that it is designed to detect defects.