# US 2013/0259376 A1 · Method for evaluating the centerline of an arbitrarily shaped object

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

- Number: US 2013/0259376 A1
- Document ID: US20130259376A1
- Status: Published application
- Kind: A1
- Application number: 13/439,778
- Filed: 2012-04-04
- Published: 2013-10-03
- Earliest claimed: 2012-03-29
- Assignees: DCG Systems, Inc.
- Inventors: Roman LOUBAN
- Web page: https://patentgrep.com/patent/US20130259376A1

## CPC

- G06T 7/66: of image moments or centre of gravity
- G06T 2207/10004: Still image; Photographic image
- G06T 2207/20021: Dividing image into blocks, subimages or windows
- G06T 2207/30152: Solder
- G06T 2207/30172: Centreline of tubular or elongated structure
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A method for calculating a centreline of an object is disclosed. An image of the object is divided into test areas. For each test area, detection direction and scanning direction are assigned from a list of limited directions. For each test area, at each scanning point a local point of the centreline is determined along the detection direction. An assigned smoothing function is applied to the collection of local points to determine the collection of pixels which define the centreline. The collection of pixels can be used to calculate the length of the centreline. Also, the coordinates of the pixels of the centreline can be used to average the intensity of the image along the centreline.

## Claims

### Claim 1

A computerized method, executed on a computer having a storage facility, for automatically determining a centerline of an object in an image, comprising:
obtaining an image of the object;
dividing the image to a plurality of test areas;
defining a limited number of scanning directions;
defining a limited number of detection directions;
for each of the plurality of test areas assigning one of the scanning directions and one of the detection directions;
examining each test area by scanning the image within the test area in the assigned scanning direction and at each scanning step performing a detection operation in the assigned detection direction to thereby designate one pixel within the detection direction as the localized centerline pixel, and thereby obtain a collection of localized centerline pixels corresponding to the centerline within the test area;
storing the coordinates of the localized centerline pixels in the storage facility; and,
summing up all the localized centerline pixels of all the test areas to thereby calculate the length of the centerline.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
defining a plurality of smoothing functions;
for each of the plurality of test areas assigning one of the smoothing functions; and,
for each of the plurality of test areas, applying the assigned smoothing function to the collection of localized centerline pixels of that test area.

### Claim 3 (depends on 2)

The method of claim 2, wherein at each examination position the localized centreline pixel is calculated as the mid-point of a line connecting two boundary points of the object on the detection direction.

### Claim 4 (depends on 2)

The method of claim 2, wherein the test areas are partially overlapping.

### Claim 5 (depends on 2)

The method of claim 2, wherein the detection direction consist essentially of the Cartesian directions: horizontal in positive x-direction, vertical in negative y-direction, 45 degrees in the first quadrant, and 45 degrees in the fourth quadrant.

### Claim 6 (depends on 5)

The method of claim 5, wherein the scanning direction is one of orthogonal or oblique to the detection direction.

### Claim 7 (depends on 2)

The method of claim 2, wherein:
the detection direction consist essentially of a horizontal direction, a vertical direction, and at least one oblique direction at 45 degrees to the horizontal direction; and,
wherein when the detection direction assigned to a test area is either the horizontal or vertical direction, the smoothing function assigned to the test area is a root mean square function and when the detection direction assigned to a test area is an oblique direction the smoothing function assigned to the test area is a morphological filtering function.

### Claim 8 (depends on 2)

The method of claim 2, further comprising calculating a difference between the length of the centerline and a reference length and indicating an error when the difference surpasses a preset threshold.

### Claim 9 (depends on 2)

The method of claim 2, further comprising calculating average intensity of all pixels belonging to the centerline.

### Claim 10

A computerized method, executed on a computer having a storage facility, for inspecting quality of a weld, comprising:
applying heating energy to the weld;
obtaining a series of images of the weld, each image taken at a defined time interval;
for each image, determining the pixels that belong to centerline of the weld;
for each image calculating average intensity of the pixels that belong to the centerline of the weld; and,
for each image storing the calculated average intensity.

### Claim 11 (depends on 10)

The method of claim 10, further comprising:
for each image, using the pixels that belong to the centerline to calculate the length of the centerline and comparing the calculated length to a reference length.

### Claim 12 (depends on 10)

The method of claim 10, wherein determining the pixels that belong to centerline comprises the steps:
dividing the image to a plurality of test areas;
defining a limited number of scanning directions;
defining a limited number of detection directions;
for each of the plurality of test areas assigning one of the scanning directions and one of the detection directions;
examining each test area by scanning the image within the test area in the assigned scanning direction and at each scanning step performing a detection operation in the assigned detection direction to thereby designate one pixel within the detection direction as the localized centerline pixel, and thereby obtain a collection of localized centerline pixels corresponding to the centerline within the test area; and,
applying a smoothing function to the collection of localized centerline pixels to thereby determine the pixels that belong to centerline.

### Claim 13 (depends on 12)

The method of claim 12, wherein the step of applying a smoothing function comprises defining a straight-line smoothing function and a curved-line smoothing function and applying the straight-line smoothing function when the detection direction is horizontal or vertical, and applying the curved-line smoothing function when the detection direction is oblique to the horizontal.

### Claim 14 (depends on 12)

The method of claim 12, wherein the scanning direction is limited to a vertical or horizontal direction, and the detection direction is at 90 degrees or 45 degrees to the scanning direction.

### Claim 15

A computerized method, executed on a computer having a storage facility, for automatically evaluating a centerline of a arbitrarily shaped object, wherein the centerline
a. is formed as a line out of interior points of the object which
b. extend along the longitudinal run of the object and
c. are each at the same distance from the next opposite side boundaries of the object, whereby
d. the length of the centerline represents the length of the object,
characterized in that the method:
a. is implemented without any change to the source code thereby that
b. the individual points of the centerline are evaluated
i. within at least one test area around the object
ii. in which the object is detected, wherein
c. as evaluation parameters
i. a pre-assigned detection direction and
ii. a pre-assigned scanning direction
are used.

### Claim 16 (depends on 15)

The method according to claim 15, characterized in that the object:
a. is examined step by step in the detection direction such that
b. the point of the centreline evaluated at each examination position is calculated as the center point of a line between the respectively detected, closest opposite side boundary points of the object whereby
c. the detection method is continued without gaps in the scanning direction
d. within the test area.

### Claim 17 (depends on 16)

The method according to claim 16 wherein the evaluation of the centerline of the object is carried out within a plurality of test areas, characterized in that the
a. the test areas are partially overlapping and
b. the centerline is composed out of portions,
c. which are formed in each test area out of the points evaluated there, whereby
d. upon composing the centerline out of the portions detected there, the overlapping of each two neighbouring test areas is taken into account.

### Claim 18 (depends on 15)

The method according to claim 15, characterized in that a detection direction is defined for each test area,
a. according to the shape and position of the portion of the object evaluated in this test area,
b. in one of two basic directions,
i. from left to right, or
ii. from top to bottom, or
c. in one of two side directions,
j. from bottom left to top right, or
jj. from top left to bottom right.

### Claim 19 (depends on 18)

The method according to claim 18, characterized in that a scanning direction is defined for each test area
a. depending on the detection direction which has already been defined, whereby
b. in case the detection direction is defined from top to bottom, the scanning direction is defined depending on the position of the neighbouring test area,
i. from left to right, or
ii. from right to left, or
c. in case the detection direction is defined as the other basic direction or one of the two side directions, the scanning direction is defined depending on the position of the neighbouring test areas,
i. from top to bottom, or
ii. from bottom to top.

### Claim 20 (depends on 15)

The method according to claim 15, characterized in that a scanning direction for each test area is evaluated depending on the previous defined detection direction as well as the position of neighbouring test areas automatically.

### Claim 21 (depends on 15)

The method according to claim 15, characterized in that the detected portions of the centerline of the test areas are smoothened individually.

### Claim 22 (depends on 21)

The method according to claim 21, characterized in that, for smoothing an evaluated portion of the center line, which originates from a straight portion of the object to be examined, a method for calculating a straight line is used.

### Claim 23 (depends on 21)

Method according to claim 21, characterized in that for smoothing an evaluated portion of the center line which originates from a non-straight portion of the object to be examined, a morphological filtering is used.
