# US 2002/0159053 A1 · Machine vision system and method for analyzing illumination lines in an image to determine characteristics of an object being inspected

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2002/0159053 A1
- Document ID: US20020159053A1
- Status: Published application
- Kind: A1
- Application number: 09/844,754
- Filed: 2001-04-27
- Published: 2002-10-31
- Earliest claimed: 2001-04-27
- Assignees: National Instruments Corporation
- Inventors: Dinesh Nair; Kevin L. Schultz
- Web page: https://patentgrep.com/patent/US20020159053A1

## CPC

- G01N 21/8806: {Specially adapted optical and illumination features}
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)

## Abstract

A machine vision system and method for performing illumination line analysis on an image of an object to detect defects in the object. The method may comprise projecting a pattern of lines on a surface of the object, and then generating an image of the surface of the object. The analysis method tracks left and right edges of each of the illumination lines to determine width and curvature of each of the lines, preferably using a bi-directional edge detection technique applied to a path perpendicular to the current orientation of the line. Information regarding the left and right edges of the line may be used to determine local widths and local orientations of the line. This information may be used to determine if a thinning or blooming of the line occurs, or if a change in curvature of the line occurs, which may indicate a possible defect in the object.

## Claims

### Claim 1

A method for analyzing an image of an object to identify defects in the object, the method comprising:
projecting a pattern of lines on a surface of the object;
generating an image of the surface of the object, wherein the image includes the pattern of lines projected on the surface of the object, wherein each line includes a left edge and right edge;
for each respective line of a plurality of the lines, repeatedly determining left and right edge points of the respective line in the image using a bi-directional edge detection technique, wherein said repeatedly determining operates to trace at least a portion of the respective line in the image;
determining whether a defect is present in the object using said left and right edge points of the plurality of the lines.

### Claim 2 (depends on 1)

The method of claim 1,
wherein said repeatedly determining comprises repeatedly determining left and right edge points of the respective line in the image using a bi-directional edge detection technique applied to a path perpendicular to the current orientation of the line.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
determining a local width of the respective line for at least a subset of the left and right edge points of the respective line; and
determining if a thinning of the respective line occurs which exceeds a threshold, wherein said determining if a thinning of the respective line occurs is based on one or more of the local widths of the respective line;
wherein a determined thinning of the respective line which exceeds the threshold indicates a possible defect in the object.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
determining a local width of the respective line for at least a subset of the left and right edge points of the respective line; and
determining if a blooming of the respective line occurs which exceeds a threshold, wherein said determining if a blooming of the respective line occurs is based on one or more of the local widths of the respective line;
wherein a determined blooming of the respective line which exceeds the threshold indicates a possible defect in the object.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
determining a local orientation of the respective line for at least a subset of the left and right edge points of the respective line; and
determining if a change in orientation of the respective line occurs which exceeds a threshold, wherein said determining if a change in orientation of the respective line occurs is based on one or more of the local orientations of the respective line;
wherein a determined change in orientation of the respective line which exceeds the threshold indicates a possible defect in the object.

### Claim 6 (depends on 1)

The method of claim 1,
wherein said repeatedly determining comprises:
repeatedly determining left and right edge points of the respective line in the image in an upward path from the at least one point;
repeatedly determining left and right edge points of the respective line in the image in a downward path from the at least one point.

### Claim 7 (depends on 1)

The method of claim 1, further comprising applying a local filter to points in the line during said repeatedly determining.

### Claim 8 (depends on 1)

The method of claim 1,
wherein said repeatedly determining left and right edge points of the respective line in the image using a bi-directional edge detection technique comprises:
at a first point in the image, determining if the first point lies within the respective line;
if the first point lies within the respective line, then performing an edge detection in a first direction away from the first point and parallel to an orientation of the respective line to determine a first edge of the respective line;
if the first point lies outside of the respective line, then performing an edge detection in a second direction away from the first point and parallel to an orientation of the respective line to determine a second edge of the respective line, wherein the second direction is opposite of the first direction.

### Claim 9 (depends on 1)

The method of claim 1,
wherein said repeatedly determining left and right edge points of the respective line in the image using a bi-directional edge detection technique comprises:
at a first point in the image, determining if the first point lies within the respective line;
if the first point lies within the respective line, then:
performing an edge detection in a first direction away from the first point and parallel to an orientation of the respective line to determine a first edge of the respective line; and
performing an edge detection in a second direction away from the first edge and parallel to an orientation of the respective line to determine a second edge of the respective line, wherein the second direction is opposite of the first direction;
if the first point lies outside of the respective line, then:
performing an edge detection in the second direction away from the first point and parallel to an orientation of the respective line to determine the second edge of the respective line; and
performing an edge detection in the first direction away from the second edge and parallel to an orientation of the respective line to determine the first edge of the respective line.

### Claim 10 (depends on 1)

The method of claim 1, wherein said repeatedly determining comprises:
determining if two or more lines overlap;
detecting when the two or more lines cease overlapping; and
continuing said repeatedly determining after the two or more lines cease overlapping.

### Claim 11

A method for tracing a line in an image, wherein the line is displayed on a surface of an object in the image, wherein the tracing is performed to identify defects in the object, wherein the line includes a left edge and a right edge, the method comprising:
a) determining left and right edge points of the line in the image, wherein said determining left and right edge points uses a bi-directional edge detection technique;
b) determining a local width of the line for the left and right edge points of the line;
c) determining a local orientation of the first line for the left and right edge points of the line;
repeating (a)-(c) a plurality of times to substantially trace the line and determine information regarding the line;
determining whether a defect is present in the object using said local widths and said local orientations.

### Claim 12 (depends on 11)

The method of claim 11,
wherein said determining left and right edge points comprises determining left and right edge points of the line in the image using a bi-directional edge detection technique applied to a path perpendicular to the current orientation of the line.

### Claim 13 (depends on 11)

The method of claim 11,
wherein said determining whether a defect is present in the object using said local widths and said local orientations comprises determining if:
1) a thinning of the line occurs which exceeds a thinning threshold, wherein a determined thinning of the line which exceeds the thinning threshold indicates a possible defect in the object;
2) a blooming of the line occurs which exceeds a blooming threshold, wherein a determined blooming of the line which exceeds the blooming threshold indicates a possible defect in the object; and/or
3) a change in orientation of the line occurs which exceeds a change in orientation threshold, wherein a determined change in orientation of the line which exceeds the change in orientation threshold indicates a possible defect in the object

### Claim 14 (depends on 11)

The method of claim 11, further comprising applying a local filter to points in the line during said determining left and right edge points of the line.

### Claim 15 (depends on 11)

The method of claim 11, further comprising:
determining if a thinning of the first line occurs based on one or more of the local widths of the first line;
determining if a change in orientation of the first line occurs based on one or more of the local orientations of the first line;
wherein a determined thinning of the first line indicates a possible defect in the object;
wherein a determined change in orientation of the first line indicates a possible defect in the object.

### Claim 16 (depends on 11)

The method of claim 11, further comprising:
projecting a pattern of lines on a surface of the object;
generating an image of the surface of the object, wherein the image includes the pattern of lines projected on the surface of the object, wherein each line includes a left edge and right edge;
wherein (a)-(c) are repeatedly performed for each of at least a subset of the pattern of lines.

### Claim 17

A method for analyzing an image of an object to identify defects in the object, the method comprising:
projecting a pattern of lines on a surface of the object;
generating an image of the surface of the object, wherein the image includes the pattern of lines projected on the surface of the object, wherein each line includes a left edge and right edge;
for each respective line of a plurality of the lines, tracking left and right edges of the respective line in the image using a bi-directional edge detection technique;
determining whether a defect is present in the object using said left and right edge points of the plurality of the lines.

### Claim 18

A method for tracing a line in an image, wherein the line is displayed on a surface of an object in the image, wherein the tracing is performed to identify defects in the object, wherein the line includes a left edge and a right edge, the method comprising:
a) tracking left and right edges of the line in the image, wherein said tracking includes determining left and right edge points uses a bi-directional edge detection technique;
b) determining a local width of the line for the left and right edge points of the line;
c) determining a local orientation of the first line for the left and right edge points of the line;
repeating steps (a)-(c) a plurality of times to substantially trace the line and determine information regarding the line;
determining whether a defect is present in the object using said local widths and said local orientations.

### Claim 19

A memory medium comprising program instructions for tracing a line in an image, wherein the line is displayed on a surface of an object in the image, wherein the tracing is performed to identify defects in the object, wherein the line includes a left edge and a right edge, wherein the program instructions are executable to implement:
a) tracking left and right edges of the line in the image, wherein said tracking includes determining left and right edge points uses a bi-directional edge detection technique;
b) determining a local width of the line for the left and right edge points of the line;
c) determining a local orientation of the first line for the left and right edge points of the line;
repeating steps (a)-(c) a plurality of times to substantially trace the line and determine information regarding the line;
determining whether a defect is present in the object using said local widths and said local orientations.

### Claim 20 (depends on 19)

The method of claim 19, further comprising applying a local filter to points in the line during said tracking left and right edges of the line, wherein the local filter operates to remove noise from pixels being analyzed.

### Claim 21

A machine vision system for analyzing an image of an object to identify defects in the object, the system comprising:
a light source configured to project a pattern of lines on a surface of the object;
a camera adapted to produce an image of the object, wherein the image includes the pattern of lines projected on the surface of the object, wherein each line includes a left edge and right edge;
a computer system coupled to the camera and adapted to receive the image of the object, wherein the computer system includes a memory medium and a processor, wherein the memory medium stores a line analysis software program, wherein, for each line of at least a subset of the pattern of lines, the processor is operable to execute the line analysis software program to:
a) determine left and right edge points of the line in the image, wherein said determining left and right edge points uses a bi-directional edge detection technique;
b) determine a local width of the line for the left and right edge points of the line;
c) determine a local orientation of the first line for the left and right edge points of the line;
repeat (a)-(c) a plurality of times to substantially trace the line and determine information regarding the line; and
determine whether a defect is present in the object using said local widths and said local orientations.

### Claim 22 (depends on 21)

The machine vision system of claim 21,
wherein the light source is configured to generate light;
wherein the machine vision system further comprises a pattern element positioned between the light source and the object, wherein the light source and the pattern element are configured to project the pattern of lines on the surface of the object.

### Claim 23 (depends on 21)

The machine vision system of claim 21, wherein the processor is operable to execute the line analysis software program to apply a local filter to points in the line during said determining left and right edge points of the line, wherein the local filter operates to remove noise from pixels being analyzed.
