Contour-based array inspection of patterned defects

1. A method of inspecting an array of cells on a substrate, the method comprising:

obtaining a first cell image that was previously determined to be defect free;

generating an array image by tiling and stitching copies of the first cell image;

generating a reference image by defining a second cell image within the array image and tiling and stitching copies of the second cell image;

generating a reference contour image which includes contours extracted from the reference image; and

using the reference contour image to detect defects in the array of cells on the substrate.

2. The method of claim 1 further comprising:

acquiring a test image of the array of cells on the substrate; and

aligning the test image to the reference contour image.

3. The method of claim 1 further comprising:

acquiring multiple overlapping higher-resolution partial images of a cell;

acquiring a lower-resolution full image of the cell; and

aligning and stitching the partial images to obtain the first cell image.

4. The method of claim 1 further comprising:

generating a cell contour image from the second cell image; and

generating the reference contour image by tiling copies of the cell contour image.

5. The method of claim 4 , wherein generating the cell contour image comprises extracting cell contours using a watershed technique.

6. The method of claim 4 , wherein generating the cell contour image comprises using polygons of the design as seeds for extracting cell contours.

7. The method of claim 4 , wherein generating the cell contour image comprises obtaining cell contours by smoothing a pre-optical-proximity-correction design.

8. The method of claim 4 , wherein generating the cell contour image comprises manually drawing cell contours using an image editing tool.

9. The method of claim 1 further comprising:

creating a mask image to provide user-specified segmentation for defect detection.

10. The method of claim 9 , wherein features are segmented out depending on a distance from edges of cell contours.

11. The method of claim 9 , wherein the user-specified segmentation is further used for defect classification.

12. A system for detecting defects in an array on a substrate, the system comprising:

a control and processing system comprising a microprocessor and control code programmed to

generate a reference image by obtaining a first cell image that was previously determined to be defect free, generating an array image by tiling and stitching copies of the first cell image, defining a second cell image within the array image, and generating a reference image by tiling and stitching copies of the second cell image,

generate a reference contour image which includes contours of extracted from the reference image, and

use the reference contour image to detect defects in the array of cells on the substrate.

13. The system of claim 12 , wherein the control code is further programmed to

acquire a test image of the array of cells on the substrate, and

align the test image to the reference contour image.

14. The system of claim 12 , wherein the control code is further programmed to

acquire multiple overlapping higher-resolution partial images of a cell,

acquire a lower-resolution full image of the cell, and

align and stitch the partial images to obtain the first cell image.

15. The system of claim 12 , wherein the control code is further programmed to

generate a cell contour image from the second cell image; and

generate the reference contour image by tiling copies of the cell contour image.

16. The system of claim 12 , wherein the control code is further programmed to

create a mask image to provide user-specified segmentation for defect detection.

17. The system of claim 16 , wherein features are segmented out depending on a distance from edges of cell contours.

18. The system of claim 16 , wherein the user-specified segmentation is further used for defect classification.