CN113418926A - Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm - Google Patents
Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm Download PDFInfo
- Publication number
- CN113418926A CN113418926A CN202110616090.7A CN202110616090A CN113418926A CN 113418926 A CN113418926 A CN 113418926A CN 202110616090 A CN202110616090 A CN 202110616090A CN 113418926 A CN113418926 A CN 113418926A
- Authority
- CN
- China
- Prior art keywords
- boundary
- contour
- point
- ith
- laser
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/12—Measuring or surveying wheel-rims
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—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
- G01N2021/8887—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 based on image processing techniques
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Mechanical Engineering (AREA)
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a wheel set tread damage detection method based on a laser strip center line positioning extraction algorithm, which relates to the technical field of railway traffic.A contour tracing extraction algorithm is utilized to extract contour boundary lines of detection laser strips from tread detection images, a center line candidate region in a preset range is screened out in the normal direction of the contour boundary lines, and a Steger algorithm is applied to the center line candidate region to position and extract the center lines of the laser strips of the detection laser strips so as to realize damage detection of the wheel set tread; the method has the advantages of small data operation amount, high extraction speed and high real-time performance, can effectively position and detect the laser stripe and eliminate the interference of stray light spots, can also effectively utilize the differential characteristic of the laser stripe image, finally realizes high-precision rapid extraction, and has good surfaces in the aspects of extraction speed and extraction precision, thereby improving the speed and precision of wheel set tread damage detection.
Description
Technical Field
The invention relates to the technical field of railway traffic, in particular to a wheel set tread damage detection method based on a laser bar center line positioning extraction algorithm.
Background
With the increasing volume of the total amount of rail transportation, the rail transportation becomes a necessary development trend of high speed and large load, which provides a severe test for detecting the safety state of the vehicle. Therefore, the quality of train parts, especially the quality of wheel sets, has a considerable influence on the operation safety of the railway transportation industry. The train has the characteristics of continuous and uninterrupted use, and each part can not avoid wearing in the continuous operation process, so that the damage of the train part needs to be detected in time, and train parts with wearing exceeding the specified limit are maintained and replaced, thereby improving the safety of the train.
The wheel set bears the weight of the whole train body, is one of the most important parts of the train, and is the most easily damaged part of the train due to continuous contact and collision with the rail in the running process of the train, so the damage detection of the wheel set is an important component of the safety detection of the train. The wheel set tread is in direct contact with a rail for a long time and inevitably becomes a main part of the wheel set which is damaged, so that how to quickly and accurately detect the damage on the wheel set tread becomes a problem to be solved urgently in the railway transportation industry, and the wheel set tread is also an important subject of long-term research of many scientific researchers in the field, and the realization of high-precision detection of the wheel set tread damage, especially on-rail automatic detection, has great significance for the development of the railway transportation industry in China.
The traditional method is that locomotive maintenance workers detect damage of locomotive wheels to treads by means of visual observation or voice recognition, and the method is low in detection efficiency and low in automation degree. With the development of machine vision detection technology, image processing technology is gradually provided to realize nondestructive detection of the locomotive wheel to the tread, for example, one method may be realized by using line-structured light scanning technology, line-structured light is projected on the surface of an object to be detected by the technology, two-dimensional center point coordinates of a distorted laser bar are extracted by the image processing technology, corresponding three-dimensional coordinates are back-calculated according to structural parameters of a line-structured light detection system, so that three-dimensional point cloud of the surface of the object is obtained, and then nondestructive detection of damage to the tread by the wheel is realized. Because the laser stripe has a certain width and has redundant information, the central line of a single-pixel width needs to be extracted, and the extraction precision of the central line directly determines the subsequent system detection precision, so how to realize a laser stripe central line extraction algorithm with high speed, high precision and good robustness is the key of the line structured light scanning technology.
The existing laser stripe center line extraction method can be mainly divided into two categories: the method mainly comprises an extreme value method, a threshold value method, a gray scale gravity center method, an edge method, a geometric center method and the like, but the methods are low in extraction accuracy and poor in robustness. And secondly, a centerline extraction method based on laser strip characteristics (such as differential characteristics, morphological characteristics and the like) mainly comprises a Steger method, a direction template method and the like, the method has high precision and certain noise resistance, but the extraction speed is still different from that of the traditional method. Therefore, the existing laser stripe center line extraction method is difficult to perform well in both extraction precision and extraction speed, and the effect of detecting wheel set tread damage by using a line structured light scanning technology is influenced.
Disclosure of Invention
Aiming at the problems and the technical requirements, the invention provides a wheel set tread damage detection method based on a laser bar center line positioning extraction algorithm, and the technical scheme of the invention is as follows:
a wheel set tread damage detection method based on a laser strip center line positioning extraction algorithm comprises the following steps:
when a locomotive runs on a track, a laser bar is projected on a wheel set tread of the locomotive through a laser, and a wheel set tread is shot through a camera to obtain a tread detection image;
extracting the contour boundary line of the detection laser stripe from the tread detection image by using a contour tracking extraction algorithm, and screening a central line region to be selected in a preset range in the normal direction of the contour boundary line;
positioning and extracting the laser bar central line of the detected laser bar by applying a Steger algorithm in the central line region to be selected;
and carrying out damage detection on the wheel set tread by utilizing a line structured light scanning technology based on the laser strip center line obtained by positioning extraction.
The further technical scheme is that the method for screening the center line candidate area in the preset range in the normal direction of the outline boundary line comprises the following steps:
sampling the contour boundary line at intervals to obtain a plurality of contour edge points;
determining a contour normal corresponding to each contour edge point by using a Hessian matrix;
and screening out the area to be selected of the central line based on the contour normal corresponding to each contour edge point.
The further technical scheme is that a contour normal corresponding to each contour edge point is determined by using a Hessian matrix, and the method comprises the following steps of:
and solving the eigenvector of the maximum eigenvalue of the Hessian matrix of the contour edge point as the normal direction corresponding to the contour edge point, and taking the line segment between the intersection point of the contour edge point and the contour boundary line in the corresponding normal direction as the contour normal corresponding to the contour edge point.
The further technical scheme is that the method for screening out the central line region to be selected based on the contour normal corresponding to each contour edge point comprises the following steps:
and selecting the area with the preset proportional width in the middle of the profile normal corresponding to each profile edge point, and screening to obtain the area to be selected of the central line.
The further technical scheme is that the method for extracting the contour boundary line of the detection laser stripe from the tread detection image by using the contour tracing extraction algorithm comprises the following steps:
extracting an interested area where the detection laser bar is located from the tread detection image according to the motion track of the wheel pair;
searching pixel points in the region of interest according to a preset searching sequence, and taking the pixel points with the searched gray value reaching a preset threshold value as boundary starting points;
and extracting contour boundary lines by using a contour tracing extraction algorithm from the boundary starting points.
The further technical scheme is that the contour tracing extraction algorithm is realized based on 8-connected chain codes, and contour boundary lines are extracted by the contour tracing extraction algorithm from a boundary starting point, and the method comprises the following steps:
determining the chain code value D of the ith boundary pointiThe corresponding chain code value of the ith boundary point in the opposite direction of the ith-1 boundary point is obtained; i is a parameter, the initial value of i is 1, and the 0 th boundary point is a boundary initial point;
if the chain code value D of the ith boundary pointiIs even number, when DiWhen-1 is not negative, will Di-1 as starting direction for the ith round of search, when DiWhen-1 is negative, (D)iThe direction corresponding to-1) +8 is used as the starting direction of the ith round of search;
if the chain code value D of the ith boundary pointiIs odd when Di-2 is a non-negative number andi-2 as starting direction for the ith round of search, when Di-2 is a negative number will (D)iThe direction corresponding to-2) +8 is used as the starting direction of the ith round of search;
with the ith boundary point as a reference point, searching an (i + 1) th boundary point of which the gray value reaches a preset threshold value from 8 adjacent points of the ith boundary point in a clockwise direction from the determined initial direction of the ith round of search;
re-executing the chain code value D for determining the ith boundary pointiAnd extracting the outline boundary line for the corresponding chain code value of the ith boundary point in the relative direction of the (i-1) th boundary point until the (i + 1) th boundary point is the boundary starting point.
The further technical scheme is that the method for extracting the outline boundary line until the (i + 1) th boundary point is taken as a boundary starting point comprises the following steps:
if the number of the extracted boundary points reaches the number threshold, extracting the contour boundary line from each found boundary point, otherwise, adjusting the search method to search each boundary point from the boundary starting point again.
The beneficial technical effects of the invention are as follows:
the application discloses a wheel set tread damage detection method based on a laser bar center line positioning extraction algorithm, the algorithm is based on a contour tracking algorithm, contour boundary lines of detection laser bars are extracted to realize the positioning of an interested area, then a center line candidate area is screened out in the normal direction of the contour boundary lines, therefore, the data operation amount is greatly reduced, the extraction speed is improved, finally, a Hessian matrix is used for extracting a center point of sub-pixel precision in the center line candidate area based on a Steger algorithm, the accurate extraction of the detection laser bars is realized, the problems that the existing extraction speed is too slow, the detection of the detection laser bars is easily interfered by stray light spots and the like are overcome, good surfaces are provided in the aspects of the extraction speed and the extraction precision, and the speed and the precision of wheel set tread damage detection can be improved.
Experimental results show that the method can realize rapid extraction of the center line of the laser bar, the consumed time is only one tenth of that of the existing Steger method, the interference of stray light spots can be effectively avoided, and high-precision center line extraction is realized.
Drawings
FIG. 1 is a flow chart of a method for detecting wheel set tread damage disclosed in the present application.
Detailed Description
The following further describes the embodiments of the present invention with reference to the drawings.
The application discloses a wheel set tread damage detection method based on a laser strip center line positioning extraction algorithm, please refer to a flow chart shown in fig. 1, and the method comprises the following steps:
step 1, in the process of running of the locomotive on a track, a laser bar is projected on a wheel set tread of the locomotive through a laser, and a wheel set tread is shot through a camera to obtain a tread detection image.
In practical application, a wheel set vision detection system is generally used for acquiring a tread detection image, the wheel set vision detection system mainly comprises a camera and a laser, the camera and the laser are respectively laid on two sides of a track, a detection laser bar is projected on a wheel set tread by the laser and a tread detection image is obtained by shooting by the camera, therefore, the acquired tread detection image comprises the wheel set tread and the detection laser bar projected on the wheel set tread, and along with the rotation of the wheel set, the tread detection image can shoot different parts of the wheel set tread and a plurality of detection laser bars projected on the wheel set tread.
And 2, extracting the contour boundary line of the detection laser stripe from the tread detection image by using a contour tracking extraction algorithm. The method specifically comprises the following substeps:
as described above, with the movement of the wheel pair, the obtained tread surface detection image may include a plurality of detection laser bars, so that the region of interest where the detection laser bar is located is extracted from the tread surface detection image according to the movement track of the wheel pair, and as the wheel pair moves linearly in the track, the detection laser bars are projected to different positions of the tread surface of the wheel pair, and the movement tracks of the detection laser bars at different positions are also linear and parallel to each other, so that the region of interest is first divided from the original tread surface detection image, and only the image region where the corresponding detection laser bar is located is covered by the region of interest, and no other detection laser bars are included, thereby avoiding the interference of the other detection laser bars. In addition, preprocessing operations such as denoising and threshold segmentation are usually required before this.
Then, searching pixel points in the region of interest according to a preset searching sequence, taking the pixel points with the searched gray value reaching a preset threshold value as boundary starting points, and searching from left to right and from top to bottom usually according to the searching sequence starting from the pixel point at the upper left corner.
And extracting contour boundary lines by using a contour tracing extraction algorithm from the boundary starting points. In one embodiment, the contour tracing extraction algorithm is implemented based on 8 connected chain codes, the chain codes are often used for representing edges of areas with similar gray feature in the fields of computer vision and the like, coding and assignment are carried out according to relative positions and directions between adjacent points, 8 adjacent points in different directions of each reference point respectively have chain code values of 0-7 in sequence, for example, the chain code value of the adjacent point right to the reference point is 0, the chain code value of the adjacent point 45 degrees below the right is 1, and the chain code value of the adjacent point right below the right is 2, and the chain code values corresponding to the critical points in different directions are sequentially obtained. The process of extracting the outline boundary line based on the 8-connected chain code is as follows:
(1) the boundary starting point is taken as the 0 th boundary point, the starting direction of the search can be predefined, for example, the right upper side is taken as the starting direction, and the pixel point of which the gray value reaches the predetermined threshold value in 8 adjacent points of the 0 th boundary point is taken as the first boundary point.
(2) Determining the chain code value D of the ith boundary pointiIs the corresponding chain code value of the ith boundary point in the opposite direction of the ith-1 th boundary point, i is a parameter and the starting value of i is 1.
(3) If the chain code value D of the ith boundary pointiIs even number, when DiWhen-1 is not negative, will Di-1 as starting direction for the ith round of search, when DiWhen-1 is negative, (D)iThe direction corresponding to-1) +8 is used as the starting direction of the ith round of search.
If the chain code value D of the ith boundary pointiIs odd when Di-2 is a non-negative number andi-2 as starting direction for the ith round of search, when Di-2 is a negative number will (D)iThe direction corresponding to-2) +8 is used as the starting direction of the ith round of search;
(4) and (3) with the ith boundary point as a reference point, searching the (i + 1) th boundary point of which the gray value reaches a preset threshold value from 8 adjacent points of the ith boundary point in a clockwise direction from the determined initial direction of the ith round of search.
(5) Re-executing the chain code value D for determining the ith boundary pointiAnd extracting the outline boundary line for the corresponding chain code value of the ith boundary point in the relative direction of the (i-1) th boundary point until the (i + 1) th boundary point is the boundary starting point.
In addition, optionally, if the number of the extracted boundary points reaches the number threshold, extracting the contour boundary line from each found boundary point, otherwise, adjusting the search method to search each boundary point from the boundary starting point again, and adjusting the search method, for example, to search in the counterclockwise direction from the starting direction, and the like, so that the interference of stray light spots can be eliminated to a certain extent, and a basis is provided for subsequently and accurately positioning the center line of the laser bar.
And 3, screening out a central line candidate area in a preset range in the normal direction of the outline boundary line. The method comprises the following specific steps:
(1) the contour boundary line is subjected to interval sampling to obtain a plurality of contour edge points, and the interference of the contour edge burr phenomenon can be reduced by adopting an interval sampling method when the normal direction is calculated.
(2) And determining the contour normal corresponding to each contour edge point by using a Hessian matrix. Specifically, the eigenvector of the maximum eigenvalue of the Hessian matrix of the contour edge point is solved as the normal direction corresponding to the contour edge point, and the line segment between the intersection point of the contour edge point and the contour boundary line in the corresponding normal direction is used as the contour normal corresponding to the contour edge point.
(3) And screening out the area to be selected of the central line based on the contour normal corresponding to each contour edge point. Specifically, selecting a region with a preset proportional width in the middle of the profile normal corresponding to each profile edge point, and screening to obtain a region to be selected of the center line. For example, if the width of the contour normal corresponding to the contour edge point is l, the middle 2l/3 width is selected as the central line region to be selected, and the method can greatly reduce the number of pixel points to be processed, so that the purpose of increasing the central line extraction speed of the laser bar is achieved.
Step 4, positioning and extracting the laser bar center line of the detected laser bar by applying a Steger algorithm in the center line candidate area, specifically: solving any point (x) of the image in the central line candidate area0,y0) The Hessian matrix can obtain the corresponding normal direction, and the unit vector of the normal direction is assumed to be (n)x,ny) Finding the image at (x)0,y0) Then find the second order Taylor expansion in the normal direction (n)x,ny) The first directional derivative of (a). Let the extracted center point coordinate of the sub-pixel be (p)x,py)＝(x0+tnx,y0+tny) Wherein (tn)x,tny) For sub-pixel offset, (p) isx,py) Substituting the first order directional derivative and making it 0 yields t, if (tn) is satisfiedx,tny)∈[-0.5,0.5]×[-0.5,0.5]Then the point with the first derivative of 0 is within the current pixel, (x)0,y0) For the extracted coordinates on the laser bar centerline at pixel level, (p)x,py) Is the corresponding sub-pixel center point coordinate. Therefore, the center line of the laser bar with sub-pixel precision can be extracted, and the accurate extraction of the center line of the laser bar is realized.
And 5, carrying out damage detection on the wheel set tread by utilizing a line structured light scanning technology based on the laser strip center line obtained by positioning extraction.
What has been described above is only a preferred embodiment of the present application, and the present invention is not limited to the above embodiment. It is to be understood that other modifications and variations directly derivable or suggested by those skilled in the art without departing from the spirit and concept of the present invention are to be considered as included within the scope of the present invention.
Claims (7)
1. A wheel set tread damage detection method based on a laser strip center line positioning extraction algorithm is characterized by comprising the following steps:
when a locomotive runs on a track, a laser bar is projected on a wheel set tread of the locomotive through a laser, and a camera is used for shooting the wheel set tread to obtain a tread detection image;
extracting contour boundary lines of the detection laser stripes from the tread detection images by using a contour tracking extraction algorithm, and screening a central line candidate region in a preset range in the normal direction of the contour boundary lines;
positioning and extracting the laser bar central line of the detection laser bar by applying a Steger algorithm in the central line region to be selected;
and carrying out damage detection on the wheel set tread by utilizing a line structured light scanning technology based on the laser strip center line obtained by positioning extraction.
2. The method of claim 1, wherein the screening out a predetermined range of candidate areas of the center line in a direction normal to the contour boundary line comprises:
sampling the contour boundary line at intervals to obtain a plurality of contour edge points;
determining a contour normal corresponding to each contour edge point by using a Hessian matrix;
and screening out the area to be selected of the central line based on the contour normal corresponding to each contour edge point.
3. The method of claim 2, wherein determining the contour normal for each contour edge point using a Hessian matrix comprises, for each contour edge point:
and solving the eigenvector of the maximum eigenvalue of the Hessian matrix of the contour edge point as the normal direction corresponding to the contour edge point, and taking a line segment between the intersection point of the contour edge point and the contour boundary line in the corresponding normal direction as the contour normal corresponding to the contour edge point.
4. The method according to claim 2, wherein the screening out the region to be selected for the center line based on the contour normal corresponding to each contour edge point comprises:
and selecting an area with a preset proportional width in the middle of the profile normal corresponding to each profile edge point, and screening to obtain the area to be selected of the central line.
5. The method according to claim 1, wherein said extracting contour boundary lines of the detected laser stripes from the tread surface detection image using a contour tracing extraction algorithm comprises:
extracting an interested area where the detection laser strip is located from the tread detection image according to the motion track of the wheel pair;
searching pixel points in the region of interest according to a preset searching sequence, and taking the pixel points with the searched gray value reaching a preset threshold value as boundary starting points;
and extracting the outline boundary line by utilizing an outline tracking extraction algorithm from the boundary starting point.
6. The method of claim 5, wherein the contour tracing extraction algorithm is implemented based on 8-connected chain codes, and the extracting the contour boundary line by using the contour tracing extraction algorithm from the boundary starting point comprises:
determining the chain code value D of the ith boundary pointiThe corresponding chain code value of the ith boundary point in the relative direction of the (i-1) th boundary point is obtained; i is a parameter, the initial value of i is 1, and the 0 th boundary point is the boundary starting point;
if the chain code value D of the ith boundary pointiIs even number, when DiWhen-1 is not negative, will Di-1 as starting direction for the ith round of search, when DiWhen-1 is negative, (D)iThe direction corresponding to-1) +8 is used as the starting direction of the ith round of search;
if the chain code value D of the ith boundary pointiIs odd when Di-2 is a non-negative number andi-2 as starting direction for the ith round of search, when Di-2 is a negative number will (D)iThe direction corresponding to-2) +8 is used as the starting direction of the ith round of search;
with the ith boundary point as a reference point, searching an i +1 th boundary point of which the gray value reaches the preset threshold value from 8 adjacent points of the ith boundary point in a clockwise direction from the determined initial direction of the ith round of search;
re-executing the chain code value D of the determined ith boundary pointiAnd extracting the outline boundary line for the corresponding chain code value of the ith boundary point in the relative direction of the (i-1) th boundary point until the (i + 1) th boundary point is the boundary starting point.
7. The method according to claim 6, wherein the extracting the contour boundary line until the i +1 th boundary point is the boundary starting point comprises:
if the number of the extracted boundary points reaches a number threshold, extracting the outline boundary line from each found boundary point, otherwise, adjusting the search method to search each boundary point from the boundary starting point again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110616090.7A CN113418926B (en) | 2021-06-02 | 2021-06-02 | Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110616090.7A CN113418926B (en) | 2021-06-02 | 2021-06-02 | Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113418926A true CN113418926A (en) | 2021-09-21 |
| CN113418926B CN113418926B (en) | 2022-02-18 |
Family
ID=77713664
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110616090.7A Active CN113418926B (en) | 2021-06-02 | 2021-06-02 | Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113418926B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102788803A (en) * | 2012-07-09 | 2012-11-21 | 南京航空航天大学 | On-line high-speed detection system of train wheel set tread and wheel rim and detection method thereof |
| CN103759671A (en) * | 2014-01-10 | 2014-04-30 | 西北农林科技大学 | Non-contact scanning method of dental cast three-dimensional surface data |
| CN104608799A (en) * | 2014-12-12 | 2015-05-13 | 郑州轻工业学院 | Information fusion technology based train wheel set tread damage online detection and recognition method |
| CN105113375A (en) * | 2015-05-15 | 2015-12-02 | 南京航空航天大学 | Pavement cracking detection system and method based on line structured light |
| CN107907048A (en) * | 2017-06-30 | 2018-04-13 | 长沙湘计海盾科技有限公司 | A kind of binocular stereo vision method for three-dimensional measurement based on line-structured light scanning |
| CN108053392A (en) * | 2017-12-05 | 2018-05-18 | 南京航空航天大学 | With reference to the wheel tread flat visible detection method of depth survey |
| CN108520521A (en) * | 2017-04-20 | 2018-09-11 | 南京航空航天大学 | The method of wheel tread extraction and splicing based on image procossing |
| CN108921864A (en) * | 2018-06-22 | 2018-11-30 | 广东工业大学 | A kind of Light stripes center extraction method and device |
| CN109389639A (en) * | 2018-07-16 | 2019-02-26 | 中国铁道科学研究院集团有限公司基础设施检测研究所 | Rail profile laser stripe center extraction method and device under dynamic environment |
| US20200166333A1 (en) * | 2016-12-07 | 2020-05-28 | Ki 'an Chishine Optoelectronics Technology Co., Ltd. | Hybrid light measurement method for measuring three-dimensional profile |
| CN111462214A (en) * | 2020-03-19 | 2020-07-28 | 南京理工大学 | Line structure light stripe central line extraction method based on Hough transformation |
| CN111784725A (en) * | 2020-06-29 | 2020-10-16 | 易思维（杭州）科技有限公司 | Light strip center extraction method |
| CN112002010A (en) * | 2020-08-12 | 2020-11-27 | 中国人民解放军总医院 | Damaged skin reconstruction method and system based on line laser scanning |
| CN112116619A (en) * | 2020-09-16 | 2020-12-22 | 昆明理工大学 | Multi-line structured light system stripe center line extraction method based on structural constraint |
| CN112489010A (en) * | 2020-11-27 | 2021-03-12 | 桂林电子科技大学 | Method and device for quickly identifying welding line and storage medium |
-
2021
- 2021-06-02 CN CN202110616090.7A patent/CN113418926B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102788803A (en) * | 2012-07-09 | 2012-11-21 | 南京航空航天大学 | On-line high-speed detection system of train wheel set tread and wheel rim and detection method thereof |
| CN103759671A (en) * | 2014-01-10 | 2014-04-30 | 西北农林科技大学 | Non-contact scanning method of dental cast three-dimensional surface data |
| CN104608799A (en) * | 2014-12-12 | 2015-05-13 | 郑州轻工业学院 | Information fusion technology based train wheel set tread damage online detection and recognition method |
| CN105113375A (en) * | 2015-05-15 | 2015-12-02 | 南京航空航天大学 | Pavement cracking detection system and method based on line structured light |
| US20200166333A1 (en) * | 2016-12-07 | 2020-05-28 | Ki 'an Chishine Optoelectronics Technology Co., Ltd. | Hybrid light measurement method for measuring three-dimensional profile |
| CN108520521A (en) * | 2017-04-20 | 2018-09-11 | 南京航空航天大学 | The method of wheel tread extraction and splicing based on image procossing |
| CN107907048A (en) * | 2017-06-30 | 2018-04-13 | 长沙湘计海盾科技有限公司 | A kind of binocular stereo vision method for three-dimensional measurement based on line-structured light scanning |
| CN108053392A (en) * | 2017-12-05 | 2018-05-18 | 南京航空航天大学 | With reference to the wheel tread flat visible detection method of depth survey |
| CN108921864A (en) * | 2018-06-22 | 2018-11-30 | 广东工业大学 | A kind of Light stripes center extraction method and device |
| CN109389639A (en) * | 2018-07-16 | 2019-02-26 | 中国铁道科学研究院集团有限公司基础设施检测研究所 | Rail profile laser stripe center extraction method and device under dynamic environment |
| CN111462214A (en) * | 2020-03-19 | 2020-07-28 | 南京理工大学 | Line structure light stripe central line extraction method based on Hough transformation |
| CN111784725A (en) * | 2020-06-29 | 2020-10-16 | 易思维（杭州）科技有限公司 | Light strip center extraction method |
| CN112002010A (en) * | 2020-08-12 | 2020-11-27 | 中国人民解放军总医院 | Damaged skin reconstruction method and system based on line laser scanning |
| CN112116619A (en) * | 2020-09-16 | 2020-12-22 | 昆明理工大学 | Multi-line structured light system stripe center line extraction method based on structural constraint |
| CN112489010A (en) * | 2020-11-27 | 2021-03-12 | 桂林电子科技大学 | Method and device for quickly identifying welding line and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| HUIPING GAO ETC.: "Robust Analysis and Laser Stripe Center Extraction for Rail Images", 《APPLIED SCIENCES》 * |
| 曾大为等: "基于自适应模板线结构的光条中心亚像素提取方法", 《光电子技术》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113418926B (en) | 2022-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107798330B (en) | Weld image feature information extraction method | |
| CN109855562B (en) | Method and device for measuring rail abrasion | |
| CN109870459B (en) | Track slab crack detection method for ballastless track | |
| CN103226834B (en) | A kind of image motion target signature point method for fast searching | |
| CN110736999B (en) | Railway turnout detection method based on laser radar | |
| CN108921164B (en) | Contact net locator gradient detection method based on three-dimensional point cloud segmentation | |
| CN111539436B (en) | Rail fastener positioning method based on straight template matching | |
| CN110189363B (en) | Airport scene moving target low-visual-angle video speed measuring method | |
| CN111260708A (en) | Line structure optical center extraction method and system | |
| CN108596872A (en) | The detection method of rail disease based on Gabor wavelet and SVM | |
| CN109653045A (en) | Gauge measurement method and device | |
| CN110490342B (en) | Contact net static geometrical parameter detection method based on Faster R-CNN | |
| CN111968079B (en) | Three-dimensional pavement crack extraction method based on local extremum of section and segmentation sparsity | |
| CN111238365B (en) | Subway train distance measurement and positioning method and system based on stereoscopic vision | |
| Wang et al. | Robust lane detection based on gradient-pairs constraint | |
| CN114119957A (en) | Method and device for detecting high-speed railway steel rail profile | |
| CN113418926B (en) | Wheel set tread damage detection method based on laser strip center line positioning extraction algorithm | |
| CN110230985B (en) | Sleeper bolt detection device | |
| CN112990192A (en) | Barcode scale image positioning, identifying and detecting method of digital level | |
| JP4459162B2 (en) | Velocity measuring device, method and program | |
| CN116681912A (en) | Rail gauge detection method and device for railway turnout | |
| CN116452826A (en) | Coal gangue contour estimation method based on machine vision under shielding condition | |
| CN116416589A (en) | Rail point cloud extraction method based on rail data and geometric properties | |
| CN112857252B (en) | Tunnel image boundary line detection method based on reflectivity intensity | |
| CN111723792B (en) | Real-time positioning point identification method suitable for rigid-flexible contact network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |