Optical pattern inspection system

• US 5,185,812 A
• Filed: 02/11/1991
• Issued: 02/09/1993
• Est. Priority Date: 02/14/1990
• Status: Expired due to Term

• Current Assignee: NuFlare Technology, Inc. (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation), Toshiba Machine Company Limited
• Inventors: Tokita, Masakazu, Yoshikawa, Ryoichi, Yamashita, Kyoji
• Primary Examiner(s): Moore, David K.
• Assistant Examiner(s): STELLRECHT, BARRY
• Application Number:
  US07/653,236
• Time in Patent Office: 729 Days
• Field of Search: 382/8, 382/18, 382/25, 356/237, 358/101, 358/106, 364/490
• US Class Current: 382/145
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06V 10/751 Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching

Abstract

A pattern inspection apparatus comprises a sensor data input section for inputting a two-dimensional inspected pattern as image data (sensor data) having a multivalued (non-binary) density distribution, design data input section for inputting reference pattern data (reference data) corresponding to the inspected pattern, a compare section for making a comparison between the image data and the reference data to obtain the difference in density therebeween, a minimum compare section for performing spatial differentiation filtering on the distribution of density between the inspected pattern and the reference pattern in different directions and obtaining the minimum of the absolute values of the filtered outputs, and a first defect determining section for detecting a defect on the basis of the minimum obtained by the compare section. In place of the first defect determining section, a maximum/minimum compare section for obtaining the minimum and maximum of the absolute values of the filtered outputs and a second defect determining section for detecting a defect on the basis of the density difference, the minimum and the maximum may be provided.

Claims

1. A pattern inspection system comprising:

first means for inputting a two-dimensional inspected pattern as image data having a multi-valued density distribution;

second means for inputting data on a reference pattern corresponding to said inspected pattern;

density difference obtaining means coupled to said first means and said second means for making a comparison between said image data and said reference data and thereby obtaining the difference in density between said inspected pattern and said reference pattern;

filter means for performing spatial differential on the distribution of the density difference between said inspected pattern and said reference pattern in different directions;

absolute value obtaining means for taking absolute values of computational results by said filter means and obtaining the minimum of the absolute values;

defect detecting means for detecting a defect of said inspected pattern on the basis of a comparison between the minimum obtained by said absolute value obtaining means and a threshold; and

defect data outputting means for outputting defect data output from said defect detecting means;

wherein said defect means detects the feature of said reference pattern data and varies the threshold for the minimum on the basis of the detected feature and determines, when the minimum exceeds its threshold, that there is a defect.

2. A pattern inspection system comprising:

first means for inputting a two-dimensional inspected pattern as image data having a multi-valued density distribution;

second means for inputting data on a reference pattern corresponding to said inspected pattern;

density difference obtaining means coupled to said first means and said second means for making a comparison between said image data and said reference data and thereby obtaining the difference in density between said inspected pattern and said reference pattern;

filter means for performing spatial differentiation on the distribution of the density difference between said inspected pattern and said reference pattern in different directions;

absolute value obtaining means for taking absolute values of computational results by said filter means and obtaining the minimum of the absolute values;

defect detecting means for detecting a defect of said inspected pattern on the basis of a comparison between the minimum obtained by said absolute value obtaining means and a threshold;

defect data outputting means for outputting defect data output from said defect detecting means; and

pattern feature extracting means which comprises storage means for storing a window region of a predetermined shape cut from design data for producing a pattern or image data obtained by taking the image of a pattern, calculating means for dividing the window region stored in said storage means into plural subregions and calculating the sum of gray levels in density of picture elements contained in each of said subregions, means for making a comparison between said subregions in the sum of gray levels and means for classifying the pattern feature of said window region on the basis of the compared results by said comparing means.

3. A pattern inspection system comprising:

first means for inputting a two-dimensional inspected pattern as image data having a multi-valued density distribution;

second means for inputting data on a reference pattern corresponding to said inspected pattern;

density difference obtaining means coupled to said first means and said second means for making a comparison between said image data and said reference data and thereby obtaining the difference in density between said inspected pattern and said reference pattern;

filter means for performing spatial differentiation on the distribution of the density difference between said inspected pattern and said reference pattern in different directions;

absolute value obtaining means for taking absolute values of computational results by said filter means and obtaining the minimum of the absolute values;

defect detecting means for detecting a defect of said inspected pattern on the basis of a comparison between the minimum obtained by said absolute value obtaining means and a threshold;

defect data outputting means for outputting defect data output from said defect detecting means; and

pattern feature extracting means which comprises storage means for storing a window region of a predetermined shape cut from design data for producing a pattern or image data obtained by taking the image of a pattern, calculating means for dividing the window region stored in said storage means into plural concentric ring-like subregions each having a constant width along the external circumference of the window region and calculating the sum of gray levels in density of picture elements contained in each of said subregions, means for obtaining the difference in the sum of gray levels between the nth (n is an integer of 1 or more) subregion and the (n+1)st subregion from the outermost subregion and means for classifying the pattern feature of the window region on the basis of the difference obtained by said means for obtaining the difference.

4. A pattern inspection system comprising:

first means for inputting a data item representing a subject pattern to be inspected;

second input means for inputting a data item representing a reference pattern which corresponds to the subject pattern;

data-comparing means connected to said first input means and said second input means, for comparing the data item representing the subject pattern, with the data item representing the reference pattern, thereby to output a data item representing a difference between the data items thus compared;

defect detecting means for detecting a defect of the data item representing said subject pattern, by processing the data item output by said data comprising means in accordance with a predetermined inspection algorithm;

memory means for storing data representing a window region cut from the subject pattern or the reference pattern and having a predetermined shape;

calculating means for dividing the window region into a plurality of sub-regions and for calculating a sum of gray levels of pixels existing in each of the sub-regions;

comparing means for comparing the sum of gray levels for each sub-region, with the sum of gray levels for any other sub-region, thereby to output data representing a difference between the sums of gray levels thus compared;

pattern-classifying means for classifying a pattern in the window region in accordance with the data output by said comparing means; and

algorithm-altering means for altering the predetermined inspection algorithm, in accordance with the feature of the pattern classified by said pattern-classifying means.

6. A pattern inspection system according to claim 4, in which each of the design data and the image data is formed of binary information comprised of 0s and 1s, and in which said calculating means for calculating a sum of gray levels calculates the number of 0s and 1s contained in each of the subregions.

5. A pattern inspection system comprising:

first input means for inputting a data item representing a subject pattern to be inspected;

second input means for inputting a data item representing a reference pattern which corresponds to the subject pattern;

data-comparing means connected to said first input means and said second input means, for comparing the data item representing the subject pattern, with the data item representing the reference pattern, thereby to output a data item representing a difference between the data item thus compared;

defect detecting means for detecting a defect of the data item representing said subject pattern, by processing the data item output by said data-comparing means in accordance with a predetermined inspection algorithm;

memory means for storing data representing a window region cut from the subject pattern or the reference pattern and having a predetermined shape;

calculating means for dividing an annular region surrounding the window region, into a plurality of concentric annular sub-regions having the same width, and for calculating a sum of gray levels of pixels existing in each of the sub-regions;

comparing means for comparing the sum of gray levels for the nth sub-region, counted from the outermost sub-region, with the sum of gray levels for the (n+1)th annular sub-region, where n is one or a greater integer, thereby to output data representing a difference between the sums of gray levels thus compared;

pattern-classifying means for classifying a pattern in the window region in accordance with the data output by said comparing means; and

algorithm-altering means for altering the predetermined detection algorithm, in accordance with the feature of the pattern classified by said pattern-classifying means.

7. A pattern inspection system according to claim 5, further comprising pattern feature extracting means for obtaining the difference in a sum of gray levels between (n+1)st and (n+2)nd subregions in addition to the difference in a sum of gray levels between the nth and (n+1)st subregions where n is an integer of 1 or more.

8. A pattern inspection system according to claim 7, further comprising pattern feature extracting means which outputs to said data comparison means data representing the feature of a pattern and based o the following processes (a), (b) and (c):

assuming the sum of gray levels of the nth (n is an integer of 1 or more) ring-like subregion to be Sn,

(a) when S_N+2 =0,

a flag is set to indicate an all-white pattern;

when S_N+2 equals the size of the (n+2)nd ring-like subregion, a flag is set to indicate an all-black pattern;

(b) when S_N -S_N+1 =S_N+1 -S_N+2,

this value itself is output;

(c) when the maximum of S_N -S_N+1 and S_N+1 -S_N+2 is 4 or below, the maximum is output;

when the minimum of S_N -S_N+1 and S_N+1 -S_N+2 is 4 or above, the minimum is output.

9. A pattern inspection system comprising:

first means for inputting a two-dimensional inspected pattern as image data having a multivalued density distribution;

second means for inputting data on a reference pattern corresponding to said inspected pattern;

density difference obtaining means, coupled to said first means and second means, for making a comparison between said image data and said reference data and thereby obtaining the difference in density between said inspected pattern and said reference pattern;

filter means for performing spatial differentiation on the distribution of the density difference between said inspected pattern and said reference pattern in different direction;

absolute value obtaining means for taking absolute values of computational results by said filter means and obtaining the minimum and maximum of the absolute values;

defect detecting means for detecting a defect on the basis of the density differences, the minimum and the maximum obtained by said absolute value obtaining means;

defect data outputting means for outputting defect data output from said defect detecting means; and

pattern feature extracting means which comprises storage means for storing a window region of a predetermined shape cut from design data for producing a pattern or image obtained by taking the image of a pattern, calculating means for dividing the window region stored in said storage means into plural subregions and calculating the sum of gray levels in density of picture elements contained in each of subregions, means for making a comparison between said subregions in the sum of gray levels and means for classifying the pattern feature of said window region on the basis of the compared results of said comparing means.

10. A pattern inspection system comprising:

first means for inputting a two-dimensional inspected pattern as image data having a multi-valued density distribution;

second means for inputting data on a reference pattern corresponding to said inspected pattern;

density difference obtaining means coupled to said first means and said second means for making a comparison between said image data and said reference data and thereby obtaining the difference in density between said inspected pattern and said reference pattern;

filter means for performing spatial differential on the distribution of the density difference between said inspected pattern and said reference pattern in different directions;

absolute value obtaining means for taking absolute values of computational results by said filter means and obtaining the minimum of the absolute values;

defect detecting means for detecting a defect of said inspected pattern on the basis of a comparison between the minimum obtained by said absolute value obtaining means and a threshold; and

defect data outputting means for outputting defect data output from said defect detecting means;

pattern feature extracting means which comprises storage means for storing a window region of a predetermined shape cut from design data for producing a pattern or image data obtained by taking the image of a pattern, calculating means for dividing the window region stored in said storage means into plural concentric ring-like subregions each having a constant width along the external circumference of the window region and calculating the sum of gray levels in density of picture elements contained in each of said subregions, means for obtaining the difference in the sum of gray levels between the nth (n is an integer of 1 or more) subregion and the (n+1)st subregion from the outermost subregion and means for classifying the pattern feature of the window region on the basis of the difference obtained by said means for obtaining the difference.