Pattern inspection method and pattern inspection device

• US 7,239,735 B2
• Filed: 12/13/2000
• Issued: 07/03/2007
• Est. Priority Date: 12/16/1999
• Status: Expired due to Fees

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Nozaki, Takeo
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): Seth; Manav
• Application Number:
  US09/735,840
• Publication Number:
  US 20010055415A1
• Time in Patent Office: 2,393 Days
• Field of Search: 382141-152, 382/266, 702 94- 95, 702 81- 84, 702190-194
• US Class Current: 382/141
• CPC Class Codes: G01N 2021/8883 involving the calculation of gauges, generating models
• CPC Class Codes: G01N 21/8851 Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges G01N21/8806 and G01N21/93 - G01N21/95692 take precedence; optical measurement of dimensions G01B11/00; optical scanning G02B26/10; image transformation G06T3/00; computerised image enhancement G06T5/00; image processing per se for flaw detection G06T7/0002
• CPC Class Codes: G01N 21/95607 using a comparative method
• CPC Class Codes: G01R 31/311 of integrated circuits G01R31/31728 takes precedence
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 7/001 using an image reference approach

Abstract

In this pattern inspection device, the optical scanning section scans the inspected pattern using a laser beam. A photoelectric image processing section generates an image of the inspected pattern. A reference image generation section calculates the gray level of each pixel according to the number of sub-pixels belonging to the pattern developed in each pixel and calculates the pattern width for the inspected pattern and the reference data with treating the count obtained by dividing the gray level by the gray level step count as the width of the pattern developed in that pixel.

Claims

1. A pattern inspection method which scans an inspected pattern formed on a substrate according to design data with a laser beam and receives light passed through said substrate with a light receiving device and, from pattern information obtained by said light receiving device, generates an image of the inspected pattern and, for coincidence between this image and reference data obtained by imaging of said design data, corrects said reference data to generate a reference image and compares the image of said inspected pattern and the reference image to detect any defects of the inspected pattern wherein,

said reference image generation comprising:

provision to each pixel of sub-pixels dividing the pixel to form a matrix and calculation of a gray level of the pixel based on the number of sub-pixels belonging to a pattern developed in each pixel,

calculation of a pattern width for the reference data by treating a number obtained by dividing said gray level by a gray level step count as the width of the pattern developed in that pixel,

determination of an edge boundary condition showing the gray level corresponding to the pattern edge position through convolution operation of an optical point spread function corresponding to a strength of the laser beam and said inspected pattern image,

detection of the edge position in said inspected pattern by the unit of sub-pixels according to said edge boundary condition,

calculation of the pattern width for said inspected pattern according to the detected edge position, and

calculation of the correction width of the pattern width for said reference data in the unit of sub-pixels from the difference between the pattern width of said inspected pattern and the pattern width of the reference data.

2. The reference image preparation method as set forth in claim 1 wherein

the gray level of each pixel in the reference data is calculated from the number of sub-pixels belonging to said reference data pattern and, treating a count obtained by dividing this gray level by the gray level step count as the pattern width of the reference data developed in the pixel,

the pattern width of the inspected pattern is calculated based on the pattern length in the unit of sub-pixels between two detected edge positions.

3. The reference image preparation method as set forth in claim 1 wherein

the pattern width for said reference data is corrected by correcting the edge position of the said reference data according to the correction width.

4. A pattern inspection device comprising:

a scanning means which scans an inspected pattern formed on the substrate according to design data with a laser beam and receives light passing through said substrate with a light receiving device;

a photoelectric image processing means which generates an image of the inspected pattern from pattern information obtained by the light receiving device in said scanning means;

a reference image generation means which generates a reference image by correcting said reference data so that positions of the image of said inspected pattern and the reference data obtained by imaging of said design data coincide;

a comparison means which compares the image of said inspected pattern and the reference image to detect any defect in the inspected pattern,

a first pattern width calculation means which provides each pixel with sub-pixels dividing the pixel into a matrix and calculates the gray level of each pixel based on the number of sub-pixels belonging to a pattern developed in each pixel and, with treating the count obtained by dividing this gray level by the gray level step count as the width of the pattern developed in the pixel, calculates the pattern width of the reference data;

an edge position detection means which determines an edge boundary condition showing the gray level corresponding to a pattern edge position through convolution operation of an optical point spread function corresponding to a strength of the laser beam strength and the image of said inspected pattern and detects an edge position of the inspected pattern by the unit of sub-pixels according to said edge boundary condition;

a second pattern width calculation means which calculates the pattern width of said inspected pattern according to the detected edge position; and

a correction width calculation means which calculates a correction width of the pattern width for said reference data in the unit of sub-pixels from the difference between the pattern width of said inspected pattern and the pattern width of the reference data.

5. The pattern inspection device as set forth in claim 4 wherein

said first pattern width calculation means

calculates the gray level of each pixel in the reference data from a number of sub-pixels belonging to the pattern of said reference data and, with treating a count obtained by dividing this gray level by the gray level step count as the pattern width of the reference data developed in the pixel, and wherein

said second pattern width calculation means calculates the pattern width of the inspected pattern based on the pattern length in the unit of sub-pixels between two detected edge positions.

6. The pattern inspection device as set forth in claim 4 wherein

said first pattern width calculation means

corrects the pattern width for said reference data by correcting the edge position of the reference data according to the correction width.

7. A computer readable memory storing a pattern inspection program which, by controlling the computer, scans the inspected pattern formed on the substrate according to the design data using the laser beam, receives the light passing from said substrate with the light receiving device, generates the image of the inspected pattern based on the pattern information obtained by the light receiving device and, for position coincidence between this image and the reference data obtained by imaging of said design data, corrects said reference data and generates the reference image, and compares the image of said inspected pattern and the reference image to detect defects of the inspected pattern wherein

said pattern inspection program executes,

in said reference image generation,

provision of sub-pixels dividing the pixel as a matrix to each pixel and calculation of a gray level for each pixel based on the number of sub-pixels belonging to a pattern developed in each pixel,

calculation of a pattern width for the reference data by treating a number obtained by dividing said gray level by a gray level step count as the width of the pattern developed in that pixel,

determination of an edge boundary condition showing the gray level corresponding to the pattern edge position through convolution operation of an optical point spread function corresponding to a strength of the laser beam and said inspected pattern image,

detection of the edge position of said inspected pattern by the unit of sub-pixels according to said edge boundary condition,

calculation of the pattern width for said inspected pattern according to the detected edge position, and

calculation of the correction width of the pattern width for said reference data in the unit of sub-pixels from the difference between the pattern width of said inspected pattern and the pattern width of the reference data.

8. The computer readable memory storing the pattern inspection program as set forth in claim 7 wherein

said pattern inspection program

calculates the gray level of each pixel from the number of sub-pixels belonging to the pattern of said reference data and, with treating a count obtained by dividing this gray level by the gray level step count as the pattern width of the reference data developed in the pixel, calculates the pattern width of said reference data, and

calculates the pattern width of the inspected pattern based on the pattern length in the unit of sub-pixels between two detected edge positions.

9. The computer readable memory storing the pattern inspection program as set forth in claim 7 wherein

said pattern inspection program

corrects the pattern width for said reference data by correcting the edge position of the said reference data according to the correction width.