# US 11,909,182 B1 · Overhead line detection method and system based on cable inspection robot

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

- Number: US 11,909,182 B1
- Document ID: US11909182B1
- Status: Granted
- Kind: B1
- Application number: 18/548,069
- Filed: 2022-11-28
- Published: 2024-03-14
- Granted: 2024-02-20
- Earliest claimed: 2022-06-20
- Assignees: State Grid Jiangsu Electric Power Co., Ltd., Changzhou Branch; State Grid Jiangsu Electric Power Co., Ltd., Innovation Center; State Grid Jiangsu Electric Power Co., Ltd.; State Grid Corporation of China
- Applicants: State Grid Jiangsu Electric Power Co., Ltd., Changzhou Branch; State Grid Jiangsu Electric Power Co., Ltd., Innovation Center; State Grid Jiangsu Electric Power Co., Ltd.; State Grid Corporation of China
- Inventors: Jiazhen Duan; Zheng Lu; Hongtao Liu; Ruxin Shi; Wei Zhang; Yuqin Shu; Zhiwei Kan; Yannan Chen; Xiaoqiang Chen; Xianming Ren
- Web page: https://patentgrep.com/patent/US11909182B1

## CPC

- H02G 1/02: for overhead lines or cables
- G01N 21/8851: {Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges (G01N21/8806 and G01N21/93 - G01N21/95692 take precedence; optical measurement of dimensions G01B11/00; optical scanning G02B26/10; image transformation G06T3/00; computerised image enhancement G06T5/00; image processing per se for flaw detection G06T7/0002)}
- G01N 21/9515: {Objects of complex shape, e.g. examined with use of a surface follower device (measuring contours and curvatures G01B11/24)}
- G01N 21/952: Inspecting the exterior surface of cylindrical bodies or wires (G01N21/956 takes precedence)
- G01N 2021/8887: {based on image processing techniques}
- G01N 2021/9518: {using a surface follower, e.g. robot}
- H02G: INSTALLATION OF ELECTRIC CABLES OR LINES, OR OF COMBINED OPTICAL AND ELECTRIC CABLES OR LINES
- 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

Provided are an overhead line detection method and system based on a cable inspection robot. The cable inspection robot is suspended on the overhead line. An imaging unit and a laser emission unit are disposed at the front end of the cable inspection robot. The imaging unit is at a first angle to the overhead line. The laser emission unit is at a second angle to the overhead line. The overhead line detection method includes the following steps: controlling the laser emission unit to emit N groups of parallel laser beams to the overhead line to form N parallel laser lines on the overhead line; controlling the imaging unit to shoot the overhead line and the N parallel laser lines to acquire a target detection image; performing image recognition on the target detection image to acquire N laser line diameters corresponding to the N parallel laser lines on the overhead line; and comparing the N laser line diameters with theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal. Therefore, the accuracy and efficiency of the inspection can be greatly improved, and the inspection risk can be reduced.

## Claims

### Claim 1

An overhead line detection method based on a cable inspection robot, wherein the cable inspection robot is suspended on an overhead line, an imaging unit and a laser emission unit are disposed at a front end of the cable inspection robot, an optical axis of the imaging unit is at a first angle to the overhead line, an optical axis of the laser emission unit is at a second angle to the overhead line, and the overhead line detection method comprises following steps:
controlling the laser emission unit to emit N groups of parallel laser beams to the overhead line to form N parallel laser lines on the overhead line, wherein N is a positive integer;
controlling the imaging unit to shoot the overhead line and the N parallel laser lines to acquire a target detection image;
performing image recognition on the target detection image to acquire N laser line diameters corresponding to the N parallel laser lines on the overhead line, wherein performing image recognition on the target detection image to acquire the N laser line diameters corresponding to the N parallel laser lines on the overhead line specifically comprises:
extracting a light strip with a width of a single pixel from a j-th parallel laser line of the N parallel laser lines by using a thinning algorithm to acquire a j-th light strip centerline, wherein 1≤j≤N, and j is a positive integer;
acquiring a pixel set of the j-th light strip centerline in the target detection image; and
performing least-square circle fitting on the pixel set to acquire a j-th laser line diameter corresponding to the j-th parallel laser line; and
comparing the N laser line diameters with theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal, wherein comparing the N laser line diameters with the theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal specifically comprises:
calculating a distances between all pixels of the j-th light strip centerline and the imaging unit by using a laser triangulation method, and selecting a first distance between a lowest pixel of the j-th light strip centerline and the imaging unit;
acquiring a fitting model, and acquiring a theoretical line diameter corresponding to the j-th laser line diameter according to the first distance by using the fitting model;
comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the j-th laser line diameter not matching the theoretical line diameter corresponding to the j-th laser line diameter, determining that the overhead line is abnormal.

### Claim 2 (depends on 1)

The overhead line detection method based on a cable inspection robot according to claim 1, wherein comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter specifically comprises:
calculating a similarity between the j-th laser line diameter and the theoretical line diameter corresponding to the j-th laser line diameter;
in response to the similarity being greater than or equal to a preset similarity, determining that the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the similarity being less than the preset similarity, determining that the j-th laser line diameter does not match the theoretical line diameter corresponding to the j-th laser line diameter.

### Claim 3 (depends on 2)

The overhead line detection method based on a cable inspection robot according to claim 2, wherein acquiring the fitting model comprises:
controlling the laser emission unit to emit K groups of parallel laser beams to an overhead line without abnormality to form K parallel laser lines on the overhead line, wherein K is a positive integer;
controlling the imaging unit to shoot the overhead line and the K parallel laser lines to acquire a sample image;
extracting a light strip with a width of a single pixel from each of the K parallel laser lines by using the thinning algorithm to acquire light strip centerlines corresponding to the K parallel laser lines;
acquiring a pixel set of each of the light strip centerlines in the sample image, and performing least-square circle fitting on the pixel set of the each of the light strip centerlines in the sample image to acquire K to-be-trained laser line diameters corresponding to the K parallel laser lines;
calculating distances between all pixels of the each of the light strip centerlines and the imaging unit by using the laser triangulation method, and selecting a second distance between a lowest pixel of the each of the light strip centerlines and the imaging unit; and
obtaining the fitting model by training a fitting algorithm according to a to-be-trained laser line diameter of the K to-be-trained laser line diameters corresponding to the each of the K parallel laser lines and the second distance.

### Claim 4

A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor, when executing the computer program, implements an overhead line detection method based on a cable inspection robot; wherein the cable inspection robot is suspended on an overhead line, an imaging unit and a laser emission unit are disposed at a front end of the cable inspection robot, an optical axis of the imaging unit is at a first angle to the overhead line, an optical axis of the laser emission unit is at a second angle to the overhead line, and the overhead line detection method comprises following steps:
controlling the laser emission unit to emit N groups of parallel laser beams to the overhead line to form N parallel laser lines on the overhead line, wherein N is a positive integer;
controlling the imaging unit to shoot the overhead line and the N parallel laser lines to acquire a target detection image;
performing image recognition on the target detection image to acquire N laser line diameters corresponding to the N parallel laser lines on the overhead line, wherein performing image recognition on the target detection image to acquire the N laser line diameters corresponding to the N parallel laser lines on the overhead line specifically comprises:
extracting a light strip with a width of a single pixel from a j-th parallel laser line of the N parallel laser lines by using a thinning algorithm to acquire a j-th light strip centerline, wherein 1≤j≤N, and j is a positive integer;
acquiring a pixel set of the j-th light strip centerline in the target detection image; and
performing least-square circle fitting on the pixel set to acquire a j-th laser line diameter corresponding to the j-th parallel laser line; and
comparing the N laser line diameters with theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal, wherein comparing the N laser line diameters with the theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal specifically comprises:
calculating a distances between all pixels of the j-th light strip centerline and the imaging unit by using a laser triangulation method, and selecting a first distance between a lowest pixel of the j-th light strip centerline and the imaging unit;
acquiring a fitting model, and acquiring a theoretical line diameter corresponding to the j-th laser line diameter according to the first distance by using the fitting model;
comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the j-th laser line diameter not matching the theoretical line diameter corresponding to the j-th laser line diameter, determining that the overhead line is abnormal.

### Claim 5

A non-transitory computer-readable storage medium storing a computer program which, when executed by a processor, implements the overhead line detection method based on a cable inspection robot;
wherein the cable inspection robot is suspended on an overhead line, an imaging unit and a laser emission unit are disposed at a front end of the cable inspection robot, an optical axis of the imaging unit is at a first angle to the overhead line, an optical axis of the laser emission unit is at a second angle to the overhead line, and the overhead line detection method comprises following steps:
controlling the laser emission unit to emit N groups of parallel laser beams to the overhead line to form N parallel laser lines on the overhead line, wherein N is a positive integer;
controlling the imaging unit to shoot the overhead line and the N parallel laser lines to acquire a target detection image;
performing image recognition on the target detection image to acquire N laser line diameters corresponding to the N parallel laser lines on the overhead line, wherein performing image recognition on the target detection image to acquire the N laser line diameters corresponding to the N parallel laser lines on the overhead line specifically comprises:
extracting a light strip with a width of a single pixel from a j-th parallel laser line of the N parallel laser lines by using a thinning algorithm to acquire a j-th light strip centerline, wherein 1≤j≤N, and j is a positive integer;
acquiring a pixel set of the j-th light strip centerline in the target detection image; and
performing least-square circle fitting on the pixel set to acquire a j-th laser line diameter corresponding to the j-th parallel laser line; and
comparing the N laser line diameters with theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal, wherein comparing the N laser line diameters with the theoretical line diameters at corresponding positions to determine whether the overhead line is abnormal specifically comprises:
calculating a distances between all pixels of the j-th light strip centerline and the imaging unit by using a laser triangulation method, and selecting a first distance between a lowest pixel of the j-th light strip centerline and the imaging unit;
acquiring a fitting model, and acquiring a theoretical line diameter corresponding to the j-th laser line diameter according to the first distance by using the fitting model;
comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the j-th laser line diameter not matching the theoretical line diameter corresponding to the j-th laser line diameter, determining that the overhead line is abnormal.

### Claim 6 (depends on 4)

The computer device according to claim 4, wherein the comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter specifically comprises:
calculating a similarity between the j-th laser line diameter and the theoretical line diameter corresponding to the j-th laser line diameter;
in response to the similarity being greater than or equal to a preset similarity, determining that the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the similarity being less than the preset similarity, determining that the j-th laser line diameter does not match the theoretical line diameter corresponding to the j-th laser line diameter.

### Claim 7 (depends on 6)

The computer device according to claim 6, wherein the acquiring the fitting model comprises:
controlling the laser emission unit to emit K groups of parallel laser beams to an overhead line without abnormality to form K parallel laser lines on the overhead line, wherein K is a positive integer;
controlling the imaging unit to shoot the overhead line and the K parallel laser lines to acquire a sample image;
extracting a light strip with a width of a single pixel from each of the K parallel laser lines by using the thinning algorithm to acquire light strip centerlines corresponding to the K parallel laser lines;
acquiring a pixel set of each of the light strip centerlines in the sample image, and performing least-square circle fitting on the pixel set of the each of the light strip centerlines in the sample image to acquire K to-be-trained laser line diameters corresponding to the K parallel laser lines;
calculating distances between all pixels of the each of the light strip centerlines and the imaging unit by using the laser triangulation method, and selecting a second distance between a lowest pixel of the each of the light strip centerlines and the imaging unit; and
obtaining the fitting model by training a fitting algorithm according to a to-be-trained laser line diameter of the K to-be-trained laser line diameters corresponding to the each of the K parallel laser lines and the second distance.

### Claim 8 (depends on 5)

The non-transitory computer-readable storage medium according to claim 5, wherein the comparing the j-th laser line diameter with the theoretical line diameter corresponding to the j-th laser line diameter to determine whether the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter specifically comprises:
calculating a similarity between the j-th laser line diameter and the theoretical line diameter corresponding to the j-th laser line diameter;
in response to the similarity being greater than or equal to a preset similarity, determining that the j-th laser line diameter matches the theoretical line diameter corresponding to the j-th laser line diameter; and
in response to the similarity being less than the preset similarity, determining that the j-th laser line diameter does not match the theoretical line diameter corresponding to the j-th laser line diameter.

### Claim 9 (depends on 8)

The non-transitory computer-readable storage medium according to claim 8, wherein the acquiring the fitting model comprises:
controlling the laser emission unit to emit K groups of parallel laser beams to an overhead line without abnormality to form K parallel laser lines on the overhead line, wherein K is a positive integer;
controlling the imaging unit to shoot the overhead line and the K parallel laser lines to acquire a sample image;
extracting a light strip with a width of a single pixel from each of the K parallel laser lines by using the thinning algorithm to acquire light strip centerlines corresponding to the K parallel laser lines;
acquiring a pixel set of each of the light strip centerlines in the sample image, and performing least-square circle fitting on the pixel set of the each of the light strip centerlines in the sample image to acquire K to-be-trained laser line diameters corresponding to the K parallel laser lines;
calculating distances between all pixels of the each of the light strip centerlines and the imaging unit by using the laser triangulation method, and selecting a second distance between a lowest pixel of the each of the light strip centerlines and the imaging unit; and
obtaining the fitting model by training a fitting algorithm according to a to-be-trained laser line diameter of the K to-be-trained laser line diameters corresponding to the each of the K parallel laser lines and the second distance.