Pattern inspecting system and pattern inspecting method

• US 6,661,507 B2
• Filed: 02/27/2002
• Issued: 12/09/2003
• Est. Priority Date: 07/04/1997
• Status: Expired due to Term

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Nozoe, Mari, Yoda, Haruo
• Primary Examiner(s): Pham, Hoa Q.
• Application Number:
  US10/083,483
• Publication Number:
  US 20020080347A1
• Time in Patent Office: 650 Days
• Field of Search: 356/237.2, 356/237.1, 356/237.3, 356/237.4, 356/237.5, 356/394, 250/307, 250/310, 250/559.06, 250/559.39, 250/559.41, 250/559.45, 250/559.46, 382/225, 382/228, 705/110, 705/28
• US Class Current: 356/237.2
• CPC Class Codes: G01N 21/956 Inspecting patterns on the surface of objects contactless testing of electronic circuits G01R31/308; testing currency G07D; manufacturing processes per se of semiconductor devices implementing a measuring step H01L22/10

Abstract

Patterns on the surface of a workpiece are inspected by calculating characteristic quantities of the patterns, classifying detected defects automatically in clusters; calculating an occurrence distribution function for each cluster, and estimating the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if the defect occurrence distribution is unknown. Some of the defects in each of the clusters are reinspected after the classifying.

Claims

1. A method of defect clustering in inspection patterns on the surface of a workpiece, said method comprising the steps of:

classifying detected defects automatically in clusters by calculated characteristic quantities of the patterns; and

estimating the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if a defect occurrence distribution is unknown.

2. A method of defect clustering in inspection patterns according to claim 1, further comprising the steps of:

(a) calculating a distance between said defects in a characteristic space;

(b) assuming that each defect forms one cluster in an initial state;

(c) setting a threshold;

(d) combining the clusters in one cluster if the distance between the nearest defects included in the clusters, is less than the threshold;

(e) increasing the threshold somewhat; and

(f) repeating from step (d) to step (e) until the threshold exceeds a predetermined upper limit or the number of the clusters in less than a predetermined lower limit.

3. A method of defect clustering in inspection patterns on the surface of a workpiece, said method comprising the steps of:

classifying detected defects automatically in clusters by calculated characteristic quantities of the patterns; and

estimating the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if a defect occurrence distribution is unknown;

reinspecting some of said defects in each said cluster.

4. A method of defect clustering in inspection patterns on the surface of a workpiece, said method comprising the steps of:

classifying detected defects automatically in clusters by calculated characteristic quantities of the patterns;

determining an occurrence probability function for each cluster on the basis of a result of said classifying step;

calculating an occurrence distribution function for each said cluster, wherein if the defect occurrence distribution is unknown, the clusters are automatically estimated from the condition of a mass of occurred defects in a characteristic space; and

clustering in the next inspection on the basis of the calculated distribution function.

5. A pattern inspecting system for inspecting patterns on the surface of a workpiece, comprising:

a calculating circuit for calculating characteristic quantities of the patterns; and

a control computer programmed to;

classify detected defects automatically in clusters;

calculate an occurrence distribution function for each of said clusters, and

estimate the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if the defect occurrence distribution is unknown.

6. A pattern inspecting system according to claim 5, wherein said control computer is further programmed to:

(a) calculate a distance between said defects in the characteristic space;

(b) assume that each defect forms one cluster in an initial state;

(c) set a threshold;

(d) combine the clusters in one cluster if the distance between the nearest defects included in the clusters, is less than the threshold;

(e) increase the threshold somewhat; and

(f) repeat from step (d) to step (e) until the threshold exceeds a predetermined upper limit or the number of the clusters in less than a predetermined lower limit.

7. A pattern inspecting system for inspecting patterns on the surface of a workpiece, comprising:

a calculating circuit for calculating characteristic quantities of the patterns; and

a control computer programmed to;

classify detected defects automatically in clusters;

calculate an occurrence distribution function for each cluster, and

estimate the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if the defect occurrence distribution is unknown;

wherein said system reinspects some of said defects in each said cluster after said classifying.

8. A pattern inspecting system for inspecting patterns on the surface of a workpiece, comprising:

a calculating circuit for calculating characteristic quantities of the patterns; and

a control computer programmed to;

classify detected defects automatically in clusters;

calculate an occurrence distribution function for each cluster on the basis of a result of said classifying, and

estimate the clusters automatically from the condition of a mass of defects that occurred in a characteristic space, if the defect occurrence distribution is unknown; and

clustering in the next inspection on the basis of the calculated distribution function.