US5544256A — Automated defect classification system

Claim 1

Independent Claim 141 words

A digital computer for automatically classifying defects in an object comprising: means for forming a defect image from a digital input image of an object; means for forming an adaptively labeled defect image from said digital input image and a reference label image of aid object; means for creating a plurality of defect feature measurements of a defect from at least said defect image, and said adaptively labeled image; fuzzy logic inference engine means having a plurality of if-then defect class rules each based on a plurality of fuzzy sets of fuzzy logic parameters, means for converting said defect feature measurements into fuzzy logic parameters and applying said converted parameters to said if-then rules to compute a delete class fuzzy set for each defect; and means lot deriving output data from each said defect class fuzzy set which classifies each defect.

Claim 2

Dependent Claim 141 words

A digital computer for automatically classifying defects in an object comprising: means for forming a defect image from a digital input image of an object; means for forming an adaptively labeled defect image from said digital input image and a reference label image of aid object; means for creating a plurality of defect feature measurements of a defect from at least said defect image, and said adaptively labeled image; fuzzy logic inference engine means having a plurality of if-then defect class rules each based on a plurality of fuzzy sets of fuzzy logic parameters, means for converting said defect feature measurements into fuzzy logic parameters and applying said converted parameters to said if-then rules to compute a delete class fuzzy set for each defect; and means lot deriving output data from each said defect class fuzzy set which classifies each defect.

Claim 3

Dependent Claim 21 words

The system of claim 1 further including means for forming said digital input image from an optical image of said object.

Claim 4

Dependent Claim 45 words

The system of claim 1 wherein said defect detecting means includes means for aligning and comparing said digital input image with a digital reference image to produce a binary defect image where ones correspond to defect pixels and all other pixels are set to zero.

Claim 5

Dependent Claim 36 words

The system of claim 3 wherein said defect detecting means includes means for labelling each pixel of said binary defect image with a label that identifies which element of said object is represented by said pixel.

Claim 6

Dependent Claim 37 words

The system of claim 4 wherein said labelling means comprises means for labelling each non-defect pixel with the same label for that pixel in a labelled digital reference image and means for adaptively labelling said defect pixels.

Claim 7

Dependent Claim 88 words

The system of claim 5 wherein said adaptive labelling means comprises means for forming a label distribution of each type of label of non-defect pixels in the vicinity of the location of defect pixel from said labelled digital reference image, means for measuring a distribution similarity parameter of said defect pixel and means for comparing the measurement of said distribution similarity parameter and said label distributions for each type of label to produce a similarity result and means for labelling said defect pixel based on said similarity result.

Claim 8

Dependent Claim 20 words

The system of claim 6 wherein said adaptive labelling means includes means for adaptively forming said labelled digital reference image.

Claim 9

Dependent Claim 20 words

The system of claim 4 wherein said feature measurement means includes means for grouping connected defect pixels into defect regions.

Claim 10

Dependent Claim 28 words

The system of claim 8 wherein said feature measurement means include means for grouping said defect region into defect clusters based on one or more region grouping criteria.

Claim 11

Dependent Claim 21 words

The system of claim 9 wherein said feature measurement means includes means for measuring and storing defect region feature measurement data.

Claim 12

Dependent Claim 27 words

The system of claim 10 wherein defect region measurement means includes means for measuring one or more of location, area, perimeter and shape of the defect regions.

Claim 13

Dependent Claim 19 words

The system of claim 11 wherein said region grouping criteria are based on said defect region feature measurement data.

Claim 14

Dependent Claim 25 words

The system of claim 12 wherein said defect regions are grouped into defect clusters based on the spatial proximity within said labelled binary defect image.

Claim 15

Dependent Claim 30 words

The system of claim 9 wherein said defect region grouping means includes means for using a union-find algorithm to compare similarities of the defect regions based on the grouping criteria.

Claim 16

Dependent Claim 21 words

The system of claim 10 wherein said feature measurement means includes means for measuring and storing defect cluster feature measurement data.

Claim 17

Dependent Claim 30 words

The system of claim 15 wherein said defect cluster feature measurement means includes means for combining the defect region feature measurement data for all defect regions within a defect cluster.

Claim 18

Dependent Claim 35 words

The system of claim 15 wherein said defect cluster feature measurement means includes means for calculating a bounding curve for all defect regions within a defect cluster measuring feature measurement data of the bounding curve.

Claim 19

Dependent Claim 32 words

The system of claim 15 wherein said defect cluster feature measurement means includes means for measuring one or more of size, shape, texture, location, composition, color and contrast of the defect cluster.

Claim 20

Dependent Claim 144 words

The system of claim 1 wherein said fuzzy logic inference engine means comprises: means for storing one or more feature fuzzy sets for each of said plurality of defect feature measurements; means for storing said plurality of if-then class rules, said rules having an output class derived from said one or more feature fuzzy sets; means for converting said defect feature measurements into fuzzy logic parameters; means for comparing said fuzzy logic parameters with each feature fuzzy set of each class rule to produce a degree of match for each feature fuzzy set for each class rule; means for combining the degree of match for each fuzzy set of each class rule to produce a degree of match for each class rule; means for combining the degree of match for each class rule to produce a defect class fuzzy set for each object defect.

Claim 21

Dependent Claim 38 words

The system of claim 19 wherein said fuzzy logic inference engine means further includes means for multiplying said degree of match for each class rule by a weighing factor to produce a scale factor for each class rule.

Claim 22

Dependent Claim 37 words

The system of claim 20 wherein said fuzzy logic inference engine means further includes means for combining the scale factor with the output class for each rule to produce an output class fuzzy set for each rule.

Claim 23

Dependent Claim 28 words

The system of claim 21 further including means for defuzzifying said defect class fuzzy set to classify the object defect as one of a plurality of defect classes.

Claim 24

Dependent Claim 17 words

The system of claim 19 wherein said plurality of if-then class rules includes supportive and exclusionary rules.

Claim 25

Dependent Claim 48 words

The system of claim 19 wherein said fuzzy logic inference engine means includes control means for producing class outputs for a subset of if-then rules based on one or more conditional expressions and means for producing said defect class fuzzy set based on the subset of class outputs.

Claim 26

Dependent Claim 17 words

The system of claim 24 wherein said conditional expressions include fuzzy operations, arithmetic operations and/or logical operations.

Claim 27

Independent Claim 118 words

A method for automatically classifying defects in an object comprising the steps of: forming a defect image from a digital input image of the object; forming an adaptively labeled defect image from said digital input image and a reference label image; creating a plurality of defect feature measurements of a defect from at least said defect image, and said adaptively labeled image; converting said defect feature measurements into fuzzy logic parameters; applying said converted parameters to a plurality of if-then rules each based on a plurality of fuzzy sets of fuzzy logic parameters; computing a defect class fuzzy set for each defect; and deriving output data from each of said defect class fuzzy sets for classifying each defect.

Claim 28

Independent Claim 118 words

A method for automatically classifying defects in an object comprising the steps of: forming a defect image from a digital input image of the object; forming an adaptively labeled defect image from said digital input image and a reference label image; creating a plurality of defect feature measurements of a defect from at least said defect image, and said adaptively labeled image; converting said defect feature measurements into fuzzy logic parameters; applying said converted parameters to a plurality of if-then rules each based on a plurality of fuzzy sets of fuzzy logic parameters; computing a defect class fuzzy set for each defect; and deriving output data from each of said defect class fuzzy sets for classifying each defect.

Claim 29

Dependent Claim 21 words

The method of claim 26 including the step of forming said digital input image from an optical image of said object.

Claim 30

Dependent Claim 43 words

The method of claim 26 wherein said defect detecting step includes aligning and comparing said digital input image with a digital reference image to produce a binary defect image where ones correspond to defect pixels and all other pixels are set to zero.

Claim 31

Dependent Claim 34 words

The method of claim 28 wherein said defect detecting step includes labelling each pixel of said binary defect image with a label that identifies which element of said object is represented by said pixel.

Claim 32

Dependent Claim 33 words

The method of claim 29 wherein said labelling step comprises labelling each non-defect pixel with the same label for that pixel in a labelled digital reference image and adaptively labelling said defect pixels.

Claim 33

Dependent Claim 80 words

The method of claim 30 wherein said adaptive labelling step comprises forming a label distribution of each type of label of non-defect pixels in the vicinity of the location of the defect pixel from said labelled digital reference image, measuring a distribution similarity parameter of said defect pixel, comparing the measurement of said distribution similarity parameter and said label distributions for each type of label to produce a similarity result, and labelling said defect pixel based on said similarity result.

Claim 34

Dependent Claim 18 words

The method of claim 31 wherein said adaptive labelling step includes adaptively forming said labelled digital reference image.

Claim 35

Dependent Claim 18 words

The method of claim 29 wherein said feature measurement step includes grouping connected defect pixels into defect regions.

Claim 36

Dependent Claim 26 words

The method of claim 33 wherein said feature measurement step includes grouping said defect region into defect clusters based on one or more region grouping criteria.

Claim 37

Dependent Claim 19 words

The method of claim 34 wherein said feature measurement step includes measuring and storing defect region feature measurement data.

Claim 38

Dependent Claim 25 words

The method of claim 35 wherein defect region measurement step includes measuring one or more of location, area, perimeter and shape of the defect regions.

Claim 39

Dependent Claim 19 words

The method of claim 36 wherein said region grouping criteria are based on said defect region feature measurement data.

Claim 40

Dependent Claim 25 words

The method of claim 37 wherein said defect regions are grouped into defect clusters based on the spatial proximity within said labelled binary defect image.

Claim 41

Dependent Claim 28 words

The method of claim 34 wherein said defect cluster grouping step includes using a union-find algorithm to compare similarities of the defect regions based on the grouping criteria.

Claim 42

Dependent Claim 19 words

The method of claim 35 wherein said feature measurement step includes measuring and storing defect cluster feature measurement data.

Claim 43

Dependent Claim 28 words

The method of claim 40 wherein said defect cluster feature measurement step includes combining the defect region feature measurement data for all defect regions within a defect cluster.

Claim 44

Dependent Claim 34 words

The method of claim 40 wherein said defect cluster feature measurement step includes calculating a bounding curve for all defect regions within a defect cluster and measuring feature measurement data of the bounding curve.

Claim 45

Dependent Claim 31 words

The method of claim 40 wherein said defect cluster feature measurement step includes measuring one or more of size, shape, texture, location, composition, color and contrast features of the defect cluster.

Claim 46

Dependent Claim 130 words

The method of claim 26 wherein said classifying step comprises: storing one or more feature fuzzy sets for each of said plurality of defect feature measurements; storing said plurality of if-then class rules, said rules having an output class derived from said one or more feature fuzzy sets; converting said defect feature measurements into fuzzy logic parameters; comparing said fuzzy logic parameters with each feature fuzzy set of each class rule to produce a degree of match for each feature fuzzy set for each class rule; combining the degree of match for each fuzzy set of each class rule to produce a degree of match for each class rule; and combining the degree of match for each class rule to produce a defect class fuzzy set for each object defect.

Claim 47

Dependent Claim 33 words

The method of claim 44 wherein said classifying step further includes multiplying said degree of match for each class rule by a weighing factor to produce a scale factor for each class rule.

Claim 48

Dependent Claim 32 words

The method of claim 45 wherein said classifying step further includes combining the scale factor with the output class for each rule to produce an output class fuzzy set for each rule.

Claim 49

Dependent Claim 29 words

The method of claim 43 further including the step of defuzzifying said defect class fuzzy set to classify the object defect as one of a plurality of defect classes.

Claim 50

Dependent Claim 17 words

The method of claim 44 wherein said plurality of if-then class rules includes supportive and exclusionary rules.

Claim 51

Dependent Claim 40 words

The method of claim 44 wherein said classifying step includes producing class outputs for a subset of if-then rules based on one or more conditional expressions and producing said defect class fuzzy sets based on the subset of class outputs.

Claim 52

Dependent Claim 17 words

The method of claim 49 wherein said conditional expressions include fuzzy operations, arithmetic operations and/or logical operations.

Claim 53

Dependent Claim 28 words

The digital computer of claim 1 wherein said plurality of defect feature measurements are created from said defect image, said adaptively labeled image and said reference label image.

Claim 54

Dependent Claim 26 words

The method of claim 26 wherein said plurality defect feature measurements are created from said defect image, said adaptively labeled image and said reference label image.

Annuity Stage	Window Opens	Payment Deadline	Surcharge Period Until	Estimated Fee	Status
3.5 Years (Year 4)	1999-08-06	2000-02-05	2000-08-06	$2000.00	Window Closed / Lapsed
7.5 Years (Year 8)	2003-08-06	2004-02-05	2004-08-06	$3760.00	Window Closed / Lapsed
11.5 Years (Year 12)	2007-08-06	2008-02-05	2008-08-06	$7700.00	Window Closed / Lapsed

Automated defect classification system

Abstract

Inventors

Bibliographic Details

Claims Analysis

References Cited (Backward Citations)

USPTO Annuity & Maintenance Windows

Assigned Categories & Tags

Assign Category Tag