Granted Patent B2

US 12,243,237 · App. 17/791,155 · Granted Mar 4, 2025

Inventor: Yosuke Okamoto (Yokohama, JP)

Assignee: TASMIT, INC.

G06T7/13G06T2207/10061G06T2207/20081G06T2207/30164

Abstract

Claims (91)

1. A pattern-edge detection method comprising:

generating an objective image of a target pattern formed on a workpiece;

generating a feature vector representing features of each pixel constituting the objective image;

inputting the feature vector to a model constructed by machine learning;

outputting, from the model, a determination result indicating whether the pixel having the feature vector is an edge pixel or a non-edge pixel; and

connecting a plurality of pixels, each having a feature vector that has obtained a determination result indicating an edge pixel, with a line to generate a virtual edge.

2. The pattern-edge detection method according to claim 1 , wherein the model is a decision tree.

3. The pattern-edge detection method according to claim 1 , further comprising:

selecting training patterns from design data;

generating training images of real patterns that have been produced based on the training patterns, respectively;

detecting edges of the real patterns on the training images;

classifying reference pixels constituting the training images into first reference pixels each constituting an edge and second reference pixels that do not constitute an edge;

generating first feature vectors of the first reference pixels and second feature vectors of the second reference pixels; and

constructing the model by the machine learning using training data including the first feature vectors and the second feature vectors.

4. The pattern-edge detection method according to claim 3 , wherein the training patterns include a plurality of patterns having at least an edge extending in a first direction, an edge extending in a second direction perpendicular to the first direction, a corner edge, and a terminal edge.

5. The pattern-edge detection method according to claim 3 , wherein the real patterns are patterns formed on the workpiece.

6. The pattern-edge detection method according to claim 3 , wherein selecting the training patterns from the design data comprises:

displaying, on a display screen, a design diagram including a plurality of patterns drawn based on the design data; and

displaying, on the display screen, the training patterns selected from the plurality of patterns included in the design diagram or an area where the training patterns are located in a visually emphasized manner.

7. The pattern-edge detection method according to claim 1 , further comprising:

generating multiple brightness profiles of the objective image along multiple search lines each extending in a normal direction with respect to the virtual edge;

determining multiple edge points based on the multiple brightness profiles; and

generating a renewed edge by connecting the multiple edge points with a line.

8. The pattern-edge detection method according to claim 7 , further comprising:

generating, from the design data, a CAD pattern corresponding to the target pattern; and

measuring a distance from an edge of the CAD pattern to the renewed edge.

9. The pattern-edge detection method according to claim 1 , further comprising:

generating a training image of a workpiece having a pattern formed thereon with a scanning electron microscope;

detecting an edge of the pattern on the training image;

calculating feature vectors of pixels constituting the training image;

dividing a target area in the training image into an edge region, a near-edge region, and a non-edge region;

producing training data including feature vectors of first pixels in the edge region, feature vectors of second pixels in the near-edge region, and feature vectors of third pixels in the non-edge region; and

producing the model by machine learning using the training data.

10. The pattern-edge detection method according to claim 9 ,

wherein where the number of first pixels is denoted by A, and a sum of the number of second pixels and the number of third pixels is denoted by B, a value (A/B) obtained by dividing the number A by the number B is a predetermined numerical value.

11. The pattern-edge detection method according to claim 10 , wherein the value (A/B) obtained by dividing the number A by the number B is in a range of 0.6 to 1.5.

12. The pattern-edge detection method according to claim 9 , wherein the non-edge region is located away from the edge region by a predetermined number of pixels, and the near-edge region is located between the edge region and the non-edge region.

13. The pattern-edge detection method according to claim 9 , wherein:

dividing the target region in the training image into the edge region, the near-edge region, and the non-edge region comprises dividing the target region in the training image into an edge region, an exclusion region, a near-edge region, and a non-edge region;

the exclusion area is adjacent to the edge region and is located between the edge region and the near-edge region; and

the training data does not include feature vectors of pixels in the exclusion region.

14. The pattern-edge detection method according to claim 9 , wherein the target region includes a first region including a first edge, a second region including a second edge perpendicular to the first edge, and a third region including a corner edge and a terminal edge.

15. The pattern-edge detection method according to claim 14 , wherein the number of pixels in the first region, the number of pixels in the second region, and the number of pixels in the third region are in a predetermined ratio.

16. A pattern-edge detection apparatus comprising:

an image generating device configured to generate an objective image of a target pattern formed on a workpiece; and

an arithmetic system coupled to the image generating device, the arithmetic system being configured to:

generate a feature vector representing features of each pixel constituting the objective image;

input the feature vector to a model constructed by machine learning;

output, from the model, a determination result indicating whether the pixel having the feature vector is an edge pixel or a non-edge pixel; and

connect a plurality of pixels, each having a feature vector that has obtained a determination result indicating an edge pixel, with a line to generate a virtual edge.

17. The pattern-edge detection apparatus according to claim 16 , wherein the model is a decision tree.

18. The pattern-edge detection apparatus according to claim 16 , wherein the arithmetic system is configured to:

select training patterns from design data;

generate training images of real patterns that have been produced based on the training patterns, respectively;

detect edges of the real patterns on the training images;

classify reference pixels constituting the training images into first reference pixels each constituting an edge and second reference pixels that do not constitute an edge;

generate first feature vectors of the first reference pixels and second feature vectors of the second reference pixels; and

construct the model by the machine learning using training data including the first feature vectors and the second feature vectors.

19. The pattern-edge detection apparatus according to claim 18 , wherein the training patterns include a plurality of patterns having at least an edge extending in a first direction, an edge extending in a second direction perpendicular to the first direction, a corner edge, and a terminal edge.

20. The pattern-edge detection apparatus according to claim 18 , wherein the real patterns are patterns formed on the workpiece.

21. The pattern-edge detection apparatus according to claim 18 , wherein the arithmetic system includes a display screen, and

the arithmetic system is configured to:

display, on the display screen, a design diagram including a plurality of patterns drawn based on the design data; and

display, on the display screen, the training patterns selected from the plurality of patterns included in the design diagram or an area where the training patterns are located in a visually emphasized manner.

22. The pattern-edge detection apparatus according to claim 16 , wherein the arithmetic system is configured to:

generate multiple brightness profiles of the objective image along multiple search lines each extending in a normal direction with respect to the virtual edge;

determine multiple edge points based on the multiple brightness profiles; and

generate a renewed edge by connecting the multiple edge points with a line.

23. The pattern-edge detection apparatus according to claim 22 , wherein the arithmetic system is configured to:

generate, from the design data, a CAD pattern corresponding to the target pattern; and

measure a distance from an edge of the CAD pattern to the renewed edge.

24. The pattern-edge detection apparatus according to claim 16 or 10 , further comprising:

a model generating apparatus, the model generating apparatus including:

a memory storing a program for producing the model; and

an arithmetic device configured to perform arithmetic operations according to instructions included in the program,

wherein the model generating apparatus is configured to:

obtain, from a scanning electron microscope, a training image of a workpiece having a pattern formed thereon;

detect an edge of the pattern on the training image;

calculate feature vectors of pixels constituting the training image;

divide a target area in the training image into an edge region, a near-edge region, and a non-edge region;

produce training data including feature vectors of first pixels in the edge region, feature vectors of second pixels in the near-edge region, and feature vectors of third pixels in the non-edge region; and

produce the model by machine learning using the training data.

25. The pattern-edge detection apparatus according to claim 24 , wherein where the number of first pixels is denoted by A, and a sum of the number of second pixels and the number of third pixels is denoted by B, a value (A/B) obtained by dividing the number A by the number B is a predetermined numerical value.

26. The pattern-edge detection apparatus according to claim 25 , wherein the value (A/B) obtained by dividing the number A by the number B is in a range of 0.6 to 1.5.

27. The pattern-edge detection apparatus according to claim 24 , wherein the non-edge region is located away from the edge region by a predetermined number of pixels, and the near-edge region is located between the edge region and the non-edge region.

28. The pattern-edge detection apparatus according to claim 24 , wherein:

the model generating apparatus is configured to divide the target region in the training image into an edge region, an exclusion region, a near-edge region, and a non-edge region;

the exclusion area is adjacent to the edge region and is located between the edge region and the near-edge region; and

the training data does not include feature vectors of pixels in the exclusion region.

29. The pattern-edge detection apparatus according to claim 24 , wherein the target region includes a first region including a first edge, a second region including a second edge perpendicular to the first edge, and a third region including a corner edge and a terminal edge.

30. The pattern-edge detection apparatus according to claim 29 , wherein the number of pixels in the first region, the number of pixels in the second region, and the number of pixels in the third region are in a predetermined ratio.

Assignments (1)

Priority Claims (2)

Continuity (1)

References Cited (19)