Apparatus and method for inspecting pattern

• US 7,689,029 B2
• Filed: 09/27/2005
• Issued: 03/30/2010
• Est. Priority Date: 09/29/2004
• Status: Expired due to Fees

• Current Assignee: SCREEN Holdings Co., Ltd.
• Original Assignee: SCREEN Holdings Co., Ltd.
• Inventors: Onishi, Hiroyuki, Ogi, Hiroshi, Asai, Hiroshi
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): Thirugnanam; Gandhi
• Application Number:
  US11/235,288
• Publication Number:
  US 20060067570A1
• Time in Patent Office: 1,645 Days
• Field of Search: 382/141, 382144-145, 382/147, 382/149
• US Class Current: 382/149
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 7/001 using an image reference approach

Abstract

An operation part in a pattern inspection apparatus includes a defect candidate image generator for generating a binary defect candidate image representing a defect candidate area in an inspection image by comparing the inspection image with a reference image, in an inspection image masking part the inspection image is masked with the defect candidate image to obtain a masked inspection image. In a feature value calculation part, an autocorrelation feature value is obtained from the masked inspection image, and outputted to a classifying part. The classifying part comprises a classifier outputting a classification result on the basis of the autocorrelation feature value and a classifier construction part for constructing the classifier by learning. It is thereby possible to easily perform the high accurate classification of defect candidate using the autocorrelation feature value which is hard to characterize as compared with geometric feature value or feature value representing density.

Claims

1. An apparatus for inspecting a pattern on an object, comprising:

a defect candidate image generator for generating a defect candidate image representing an area which includes a defect candidate in a grayscale inspection image representing pattern on an object by comparing said inspection image with a reference image;

an inspection image masking part for masking said inspection image with said defect candidate image to obtain a masked inspection image;

a reference image masking part for masking said reference image with said defect candidate image to obtain a masked reference image;

a feature value calculation part for obtaining an autocorrelation feature value from each of said masked inspection image and said masked reference image, each of said autocorrelation feature values being obtained as a high-dimensional vector of which values of elements are obtained by using a plurality of autocorrelation matrices; and

a classifying part for performing a classification of said defect candidate on the basis of said autocorrelation feature values, wherein

said classifying part comprises a classifier construction part for constructing a classifier by learning which outputs a classification result in accordance with said autocorrelation feature values.

2. The apparatus according to claim 1, wherein

said classification is also based on an autocorrelation feature value obtained from said defect candidate image.

3. A method for inspecting a pattern on an object, comprising the steps of:

a) generating a defect candidate image representing an area which includes a defect candidate in a grayscale inspection image representing pattern on an object by comparing said inspection image with a reference image;

b) masking said inspection image with said defect candidate image to obtain a masked inspection image;

c) masking said reference image with said defect candidate image to obtain a masked reference image;

d) obtaining an autocorrelation feature value from each of said masked inspection image and said masked reference image, each of said autocorrelation feature values being obtained as a high-dimensional vector of which values of elements are obtained by using a plurality of autocorrelation matrices; and

e) performing a classification of said defect candidate on the basis of said autocorrelation feature values, wherein

said classification is performed by a classifier constructed by learning which outputs a classification result in accordance with said autocorrelation feature values, and

the above steps are performed by a computer.

4. The method according to claim 3, wherein

said classification is also based on an autocorrelation feature value obtained from said defect candidate image.