Pattern inspection system

1. A pattern inspection system which inspects an image of an inspection target pattern of an electronic device using an identifier constituted by machine learning, based on the image of the inspection target pattern of the electronic device and data used to manufacture the inspection target pattern, the system comprising:

a storage unit which stores a plurality of pattern images of the electronic device and pattern data used to manufacture a pattern of the electronic device; and

an image selection unit which selects a learning pattern image used in the machine learning from the plurality of pattern images, based on the pattern data and the pattern image stored in the storage unit; and

a region extraction unit which extracts one of a contour shape of at least a pattern, a pattern region, and a non-pattern region from the image of the inspection target pattern by the identifier.

2. The pattern inspection system according to claim 1 , further comprising:

an identifier generation unit which executes machine learning based on the learning pattern image and a true value data corresponding to the learning pattern image so as to generate the identifier.

3. The pattern inspection system according to claim 2 , wherein

the identifier extracts a defect based on the image of the inspection target pattern.

4. The pattern inspection system according to claim 3 , wherein

the identifier generation unit creates, from the learning pattern image, teaching data tagged with information of a type of defect for each image by a user, and executes learning using the teaching data.

5. The pattern inspection system according to claim 2 , wherein

identification data obtained by the identifier and the pattern data used to manufacture the pattern of the electronic device are compared with each other so as to execute inspection.

6. The pattern inspection system according to claim 2 , wherein

the identifier generation unit has a display unit for displaying the learning pattern image, and an instruction unit for enabling a user to execute a true value assignment to the learning pattern image displayed on the display unit, and executes learning using true value assigned teaching data.

7. The pattern inspection system according to claim 2 , further comprising:

a model evaluation unit which evaluates the image of the inspection target pattern using a model which is generated in the identifier generation unit by machine learning based on the learning pattern image and the true value data corresponding to the learning pattern image.

8. The pattern inspection system according to claim 2 , wherein

the identifier executes a region division of the image and classification of the image.

9. The pattern inspection system according to claim 1 , wherein

the image selection unit executes at least one of analyzing a shape of a pattern using the pattern data, analyzing a position of a pattern on a semiconductor device, and analyzing a variation of the image using the pattern image.

10. The pattern inspection system according to claim 9 , wherein

the image selection unit, as processing of analyzing the image variation, executes statistical processing using the pattern image, using at least one of a value of a variation of a white band width, a value of a variation of roughness, and a value of a variation of luminance values of a pattern region and a non-pattern region.

11. The pattern inspection system according to claim 1 , wherein the identifier generation unit divides the learning pattern image into a contour line of a pattern, a pattern region, and a non-pattern region at each of a single layer and multilayer of a semiconductor device by a user, and executes learning using true value assigned teaching data for each pixel.

12. A pattern inspection system which inspects an image of an inspection target pattern of an electronic device using an identifier constituted by machine learning, based on the image of the inspection target pattern of the electronic device and data used to manufacture the inspection target pattern, the system comprising:

a storage unit which stores pattern data used to manufacture a pattern of the electronic device and photographing condition data of the image of the inspection target pattern;

a photographing position selection unit which selects a photographing position of a learning pattern image on the electronic device used in the machine learning, based on the pattern data and the photographing condition data stored in the storage unit; and

a region extraction unit which extracts one of a contour shape of at least a pattern, a pattern region, and a non-pattern region from the image of the inspection target pattern by the identifier.

13. The pattern inspection system according to claim 12 , wherein

the photographing position selection unit analyzes a shape of a pattern using the pattern data and/or analyzes a position of a pattern on a semiconductor device.

14. The pattern inspection system according to claim 13 , wherein

the image selection unit executes statistical processing of the number of pixels of a vertical edge and a horizontal edge based on edge information of a pattern obtained pattern data corresponding to the pattern image to derive the same shape.

15. The pattern inspection system according to claim 13 , wherein

the image selection unit detects a position on the pattern data in which the shape of the same pattern exists, using a plurality of patterns having the same shape detected by analyzing a shape of the pattern and photographing condition data of the pattern image, and executes statistical processing using at least one of the detected number, a coordinate position on a semiconductor chip, a coordinate position on a wafer, and distance information between the coordinate position on the semiconductor chip and the coordinate position on the wafer, and the detected position.

16. The pattern inspection system according to claim 12 , further comprising:

a recipe creation unit which generates a recipe of a photographing unit based on a photographing position selected by the photographing position selection unit.

17. The pattern inspection system according to claim 12 , wherein

the photographing position selection unit divides the pattern data into small areas corresponding to an imaging field of view of the photographing condition data, analyzes a shape of a pattern for each divided area, and analyzes a position of the pattern on the electronic device based on a pattern having the same shape as the shape of the analyzed pattern to specify the photographing position information, the photographing condition, and additional information from a user.

18. A pattern inspection system which inspects an image of an inspection target pattern of an electronic device using an identifier constituted by machine learning, based on the image of the inspection target pattern of the electronic device and data used to manufacture the inspection target pattern, the system comprising:

a storage unit which stores pattern data used to manufacture a pattern image of the electronic device and a pattern of the electronic device and photographing condition data of the image of the inspection target pattern;

an image selection unit which selects a learning pattern image used in the machine learning, based on the pattern data, the pattern image, and the photographing condition data stored in the storage unit; and

a region extraction unit which extracts one of a contour shape of at least a pattern, a pattern region, and a non-pattern region from the image of the inspection target pattern by the identifier.

19. The pattern inspection system according to claim 18 , wherein

the image selection unit executes at least one of analyzing a shape of a pattern using the pattern data, analyzing a position of a pattern on a semiconductor device, and analyzing a variation of the image using the pattern image.