IP Library Granted Patent US 7,796,803
Granted Patent B2
US 7,796,803 · App. 11/567,550 · Granted Sep 14, 2010

Image correction method and apparatus for use in pattern inspection system

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,796,803
App. No.
11/567,550
Granted
Sep 14, 2010
Kind
B2
Abstract

A technique for correcting an image by using frequency division images and decomposition images corresponding in number to reference points is disclosed. An image correction apparatus includes an image divider which divides an inspection reference image into frequency regions to form frequency division images, a decomposition image generator for defining reference points at several locations within at least one frequency division image and for applying weighting with each reference point as a reference to thereby generate decomposition images corresponding in number to the reference points, a model parameter identifier for identifying a model parameter by using 2D linear prediction models of an image being tested, the decomposition images and a frequency division image which is out of the generation of decomposition images, and a model image generator for using the model parameter to generate a model image. An image inspection apparatus using the correction apparatus and an image correction method are also disclosed.

Claims (20)

1. An image correction apparatus for generating a model image from an inspection reference image and an image to be tested, said apparatus comprising:

an image division unit operative to divide the reference image into a plurality of frequency regions to thereby form a plurality of frequency division images;

a decomposition image generation unit operative to provide reference points at a plurality of spaced-apart locations within at least one of the frequency division images and apply weighting to the frequency division image with each reference point as a reference to thereby generate decomposition images corresponding in number to the reference points;

a model parameter identification unit operative to identify a model parameter by use of two-dimensional linear prediction models of the to-be-tested image, the decomposition images corresponding in number to the reference points and a frequency division image which does not generate decomposition images; and

a model image generation unit operative to generate a model image by using the model parameter identified.

2. The apparatus according to claim 1 , wherein said model parameter identification unit identifies the model parameter from a set of simultaneous equations describing therein an input/output relationship using the two-dimensional linear prediction models regarding each pixel of the to-be-tested image as an output and letting a linear coupling of pixels around each pixel be an input with respect to the decomposition images corresponding in number to the reference points and the frequency division image which does not generate decomposition images.

3. The apparatus according to claim 1 , wherein said image division unit divides the reference image into a first frequency region and a second frequency region, the second frequency region being higher in frequency than the first region, to thereby form two frequency division images,

said decomposition image generation unit provides the reference points at respective apexes of a frequency division image of the first frequency region and applies weighting to the frequency division image with each reference point as a reference to thereby generate decomposition images corresponding in number to the reference points, and

said model parameter identification unit identifies the model parameter from the simultaneous equations describing the input/output relationship using the two-dimensional linear prediction model regarding each pixel of the to-be-tested image as an output and letting the linear coupling of pixels around each pixel be an input with respect to the decomposition images corresponding in number to the reference points and a reference frequency division image which does not generate decomposition images.

4. An image inspection apparatus for inspecting through comparison an inspection reference image and an image being tested, said apparatus comprising:

an image division unit operative to divide the reference image into a plurality of frequency regions to thereby form a plurality of frequency division images;

a decomposition image generation unit operative to provide reference points at a plurality of spaced-apart locations within at least one of the frequency division images and apply weighting to the frequency division image with each reference point as a reference to thereby generate decomposition images corresponding in number to the reference points;

a model parameter identification unit operative to identify a model parameter by use of two-dimensional linear prediction models of the test image, the decomposition images corresponding in number to the reference points and a frequency division image which does not generate decomposition images;

a model image generation unit operative to generate a model image by using the model parameter identified; and

a comparison processing unit for performing comparison inspection of the model image and the test image.

5. An image correction method, performed by a processor or computer, for generating a model image from an inspection reference image and an image to be tested, said method comprising:

dividing, at the processor or computer, the reference image into a plurality of frequency regions to thereby form a plurality of frequency division images;

providing, at the processor or computer, reference points at a plurality of spaced-apart locations within at least one of the frequency division images and applying weighting to the frequency division image with each reference point as a reference to thereby generate decomposition images corresponding in number to the reference points;

identifying, at the processor or computer, a model parameter by use of two-dimensional linear prediction models of the to-be-tested image, the decomposition images corresponding in number to the reference points, and a frequency division image which does not generate decomposition images; and

generating, at the processor or computer, a model image by using the model parameter identified.

Assignments (6)
MERGER AND CHANGE OF NAME Recorded Aug 19, 2021
From: TOSHIBA MEMORY CORPORATION; K.K. PANGEA
To: TOSHIBA MEMORY CORPORATION
Reel/Frame 057224/0091 →
CHANGE OF NAME Recorded Aug 19, 2021
From: TOSHIBA MEMORY CORPORATION
To: KIOXIA CORPORATION
Reel/Frame 057224/0641 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 3, 2018
From: KABUSHIKI KAISHA TOSHIBA
To: TOSHIBA MEMORY CORPORATION
Reel/Frame 045574/0652 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 20, 2010
From: ADVANCED MASK INSPECTION TECHNOLOGY INC.
To: KABUSHIKI KAISHA TOSHIBA; NEC CORPORATION
Reel/Frame 025008/0164 →
CORPORATE ADDRESS CHANGE Recorded Jun 5, 2007
From: ADVANCED MASK INSPECTION TECHNOLOGY INC.
To: ADVANCED MASK INSPECTION TECHNOLOGY INC.
Reel/Frame 019385/0760 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 6, 2006
From: OAKI, JUNJI
To: ADVANCED MASK INSPECTION TECHNOLOGY INC.
Reel/Frame 018592/0065 →