CN111144235B - Switch tongue track crawling monitoring method based on video - Google Patents
Switch tongue track crawling monitoring method based on video Download PDFInfo
- Publication number
- CN111144235B CN111144235B CN201911258437.4A CN201911258437A CN111144235B CN 111144235 B CN111144235 B CN 111144235B CN 201911258437 A CN201911258437 A CN 201911258437A CN 111144235 B CN111144235 B CN 111144235B
- Authority
- CN
- China
- Prior art keywords
- point
- switch
- rail
- edge
- climbing
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a video-based switch blade crawling monitoring method, which is used for solving the problem that the existing switch blade crawling monitoring is low in monitoring efficiency and monitoring accuracy, and comprises the following steps: setting high-definition cameras on two sides of the switch rail and calibrating the positions of the cameras; marking tag points on a stock rail by manpower; step two: acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; step three: the climbing monitoring is realized through the recognition of the edge outline of the fork tip; the climbing includes a longitudinal climbing of the switch rail along the extending direction of the stock rail; according to the invention, the high-definition cameras are arranged on the two sides of the switch rail, the actual distance between the switch rail and the stock rail is calculated by utilizing the edge identification image of the switch point, and the accuracy of monitoring is improved by combining the identification of the edge profile of the switch point.
Description
Technical Field
The invention relates to the technical field of switch blade crawling monitoring, in particular to a switch blade crawling monitoring method based on video.
Background
Two general types of faults affecting operational safety occur in switches and their point machine equipment. One is a switch machine, and the action of the switch machine is completed by sending out current and voltage control signals indoors, but once the current and voltage work abnormally, an action rod moves transversely or a rail crawls, etc., the switch cannot be switched according to the set requirements, so that accidents occur. The other is a turnout point, because the trend of the train is controlled by the clutch between the point rail and the stock rail, the longitudinal climbing of the point rail and the stock rail and the damage of the point rail have the danger of derailing the train, the abrasion of the high-speed rail point rail is very quick, the climbing effect obviously causes the abnormal action of the switch machine, so the abrasion of the point rail, the climbing of the point rail, whether the point rail is closed with the stock rail (accidents are caused when the gap exceeds 2-4 mm) and the like are factors influencing the operation safety;
aiming at the two types of faults, the mode of manual inspection and equipment detection is mainly adopted in China at present. Except for the manual inspection mode, the domestic equipment for solving the two problems is an independent system. The phenomenon of rail crawling or action rod traversing can cause abnormal work of the switch machine. Therefore, a video-based switch tongue crawling monitoring method is needed to be designed mainly aiming at the disasters which are easy to occur;
in recent years, with the development of CMOS camera processes and the improvement of hardware computing speed, image-based target state monitoring is gradually replacing the conventional monitoring method. Especially after the traditional machine vision is introduced into a learning algorithm, the accuracy of a novel switch blade crawling monitoring method based on video is continuously improved, and compared with the existing monitoring method, the monitoring method based on video/image can directly reflect the switch blade state, and can be matched with a corresponding target detection algorithm, so that the real-time monitoring of switch blade crawling monitoring can be realized, and the monitoring efficiency and accuracy can be greatly improved.
Disclosure of Invention
The invention aims to solve the problems of low monitoring efficiency and monitoring accuracy of the existing switch point close contact degree monitoring method, and provides a switch point crawling monitoring method based on video; according to the invention, the high-definition cameras are arranged on the two sides of the switch rail, the actual distance between the switch rail and the stock rail is calculated by utilizing the edge identification image of the switch point, and the accuracy of monitoring is improved by combining the identification of the edge profile of the switch point.
The aim of the invention can be achieved by the following technical scheme: a video-based switch tongue crawling monitoring method, the method comprising the steps of:
step one: setting high-definition cameras on two sides of the switch rail and calibrating the positions of the cameras; marking tag points on a stock rail by manpower;
step two: acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; the edge recognition of the turnout point comprises the edge recognition of the stock rail and the edge recognition that the turnout point part is contacted with the stock rail;
the edge recognition of the fork point comprises the following steps:
s1: determining the position of a stock rail through manually calibrating a tag point;
s2: acquiring a binary image containing target edge information from the acquired image through a Canny edge algorithm, then acquiring a discrete coordinate point by utilizing Hough transformation and edge detection, and extracting continuous edge information to acquire a stock rail and a fork point edge;
s3: after identifying the stock rail and the fork tip edge, calculating to obtain the distance between the switch rail and the stock rail;
s4: the actual distance is restored to obtain the actual distance between the switch rail and the stock rail through the scale invariance of the image recognition target distance in the same direction;
step three: the climbing monitoring is realized through the recognition of the edge outline of the fork tip; the climbing includes a longitudinal climbing of the switch rail along the extending direction of the stock rail; longitudinal climbing includes both climbing of the point rail relative to the stock rail and relative climbing of the two point rails; longitudinal crawler monitoring includes identification based on reference points that can indicate the location of the fork point and identification based on the edge profile of the fork point.
Preferably, the step of identifying the edge profile of the fork tip in the step three includes the following steps:
SS1: the high-definition cameras arranged on the two sides of the switch rail are respectively marked as a camera A and a camera B; the point where the camera A is located is the point A, and the point where the camera B is located is the point B;
SS2: taking the position of the camera B as a reference zero point and taking the vertical position as a Z axis; establishing a three-dimensional coordinate system;
SS3: the identified fork tip edge obtains two reference points for indicating the position of the fork tip; a connection of two reference points indicating the position of the fork point; the fork tip is monitored for climbing through the connecting line parallel to the X axis.
Preferably, the step three of identifying the edge profile based on the fork tip further includes the following steps:
SS1: manually attaching a plurality of label points on one side of the switch rail;
SS2: photographing the tag point by a high-definition camera at one side of the switch rail; and marks it as an initial tag point, denoted by APi, i=1, 2, … …, n;
SS3: when the switch point is monitored, the tag point is photographed again and marked as a real-time tag point, and the real-time tag point is denoted by BPi;
SS4: matching the real-time tag points with the initial tag points one by one; when it is deviated, it means that the point rail climbs longitudinally in the direction of the stock rail.
Compared with the prior art, the invention has the beneficial effects that:
setting high-definition cameras on two sides of the switch rail and calibrating the positions of the cameras; marking tag points on a stock rail by manpower; acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; the edge recognition of the turnout point comprises the edge recognition of the stock rail and the edge recognition that the turnout point part is contacted with the stock rail; the climbing monitoring is realized by utilizing the recognition of the edge contour of the fork tip; by arranging high-definition cameras on two sides of the switch rail, the actual distance between the switch rail and the stock rail is calculated by utilizing the identification images of the edge contours of the switch points, and the accuracy of monitoring is improved by combining the identification of the edge contours of the switch points.
Drawings
The present invention is further described below with reference to the accompanying drawings for the convenience of understanding by those skilled in the art.
Fig. 1 is a schematic view of a reference point based on the point position of the switch point.
FIG. 2 is a schematic illustration of the present invention based on recognition of a fork tip edge profile.
FIG. 3 is a schematic illustration of another embodiment of the present invention based on the identification of a fork tip edge profile.
Detailed Description
The technical solutions of the present invention will be clearly and completely described in connection with the embodiments, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-3, a video-based switch blade crawling monitoring method, the video-based switch blade crawling monitoring method comprises the following steps:
step one: setting high-definition cameras on two sides of the switch rail and calibrating the positions of the cameras; marking tag points on a stock rail by manpower; marking the tag points as key components, and detecting the key components such as the switch rail bolts by using a corresponding target recognition algorithm;
step two: acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; the edge recognition of the turnout point comprises the edge recognition of the stock rail and the edge recognition that the turnout point part is contacted with the stock rail;
the edge recognition of the fork point comprises the following steps:
s1: determining the position of a stock rail through manually calibrating a tag point;
s2: acquiring a binary image containing target edge information from the acquired image through a Canny edge algorithm, then acquiring a discrete coordinate point by utilizing Hough transformation and edge detection, and extracting continuous edge information to acquire a stock rail and a fork point edge;
s3: after identifying the stock rail and the fork tip edge, calculating to obtain the distance between the switch rail and the stock rail; the specific calculation steps are as follows:
s4: the actual distance is restored to obtain the actual distance between the switch rail and the stock rail through the scale invariance of the image recognition target distance in the same direction;
as shown in fig. 1, denoted by x1 as the distance between the point rail and the stock rail farther away; taking a camera A as a reference point, wherein x1 calculated by the camera A is the distance between a stock rail close to the camera A side and a point rail close to the camera B side;
using tangent functions
The length of x1 can be calculated, by the length of x 1; setting the distance between the camera and the vertical line to be beta, and measuring the actual distance between the switch rail and the stock rail through the camera A and the camera B, wherein alpha=90-beta;
step three: the climbing monitoring is realized through the recognition of the edge outline of the fork tip; the climbing includes a longitudinal climbing of the switch rail along the extending direction of the stock rail; longitudinal climbing includes both climbing of the point rail relative to the stock rail and relative climbing of the two point rails; the longitudinal crawler monitoring comprises identification based on a reference point capable of indicating the position of the fork point and identification based on the edge profile of the fork point;
example 1, the recognition procedure for the fork tip edge profile recognition shown in fig. 2 is as follows:
SS1: the high-definition cameras arranged on the two sides of the switch rail are respectively marked as a camera A and a camera B; the point where the camera A is located is the point A, and the point where the camera B is located is the point B;
SS2: taking the position of the camera B as a reference zero point and taking the vertical position as a Z axis; establishing a three-dimensional coordinate system;
SS3: the identified fork tip edge obtains two reference points for indicating the position of the fork tip; a line of two reference points indicating the positions of the fork points, namely a line segment cd; the line cd is parallel to the X axis and is used for monitoring the climbing movement of the fork tip;
example 2, as shown in fig. 3, the fork tip edge profile identification further includes the following steps:
SS1: manually attaching a plurality of label points on one side of the switch rail;
SS2: photographing the tag point by a high-definition camera at one side of the switch rail; and marks it as an initial tag point, denoted by APi, i=1, 2, … …, n;
SS3: when the switch point is monitored, the tag point is photographed again and marked as a real-time tag point, and the real-time tag point is denoted by BPi;
SS4: matching the real-time tag points with the initial tag points one by one; when the deviation occurs, the longitudinal climbing of the switch rail along the direction of the stock rail is indicated; the specific matching process is as follows:
establishing a rectangular coordinate system for the shot tag point images, respectively counting the coordinates of the real-time tag points, and judging that deviation occurs when the coordinates of the real-time tag are different from the coordinates of the corresponding initial tag points;
when the high-definition camera is used, high-definition cameras are arranged on two sides of the switch rail, and the positions of the cameras are calibrated; marking tag points on a stock rail by manpower; acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; the edge recognition of the turnout point comprises the edge recognition of the stock rail and the edge recognition that the turnout point part is contacted with the stock rail; the climbing monitoring is realized by utilizing the recognition of the edge contour of the fork tip; by arranging high-definition cameras on two sides of the switch rail, the actual distance between the switch rail and the stock rail is calculated by utilizing the fork point edge profile identification image, and the accuracy of monitoring is improved by combining the identification of the fork point edge profile.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (3)
1. The track switch blade crawling monitoring method based on the video is characterized by comprising the following steps of:
step one: setting high-definition cameras on two sides of the switch rail and calibrating the positions of the cameras; marking tag points on a stock rail by manpower;
step two: acquiring videos of the switch rails through two high-definition cameras, dividing the videos into images, and identifying the images by utilizing the edges of the switch points; the edge recognition of the turnout point comprises the edge recognition of the stock rail and the edge recognition that the turnout point part is contacted with the stock rail;
the edge recognition of the fork point comprises the following steps:
s1: determining the position of a stock rail through manually calibrating a tag point;
s2: acquiring a binary image containing target edge information from the acquired image through a Canny edge algorithm, then acquiring a discrete coordinate point by utilizing Hough transformation and edge detection, and extracting continuous edge information to acquire a stock rail and a fork point edge;
s3: after identifying the stock rail and the fork tip edge, calculating to obtain the distance between the switch rail and the stock rail;
s4: the actual distance is restored to obtain the actual distance between the switch rail and the stock rail through the scale invariance of the image recognition target distance in the same direction;
step three: the climbing monitoring is realized through the recognition of the edge outline of the fork tip; the climbing includes a longitudinal climbing of the switch rail along the extending direction of the stock rail; longitudinal climbing includes both climbing of the point rail relative to the stock rail and relative climbing of the two point rails; longitudinal crawler monitoring includes identification based on reference points that can indicate the location of the fork point and identification based on the edge profile of the fork point.
2. The method of video-based switch blade crawling monitoring according to claim 1, wherein the step of identifying the edge profile of the switch blade in step three is as follows:
SS1: the high-definition cameras arranged on the two sides of the switch rail are respectively marked as a camera A and a camera B; the point where the camera A is located is the point A, and the point where the camera B is located is the point B;
SS2: taking the position of the camera B as a reference zero point and taking the vertical position as a Z axis; establishing a three-dimensional coordinate system;
SS3: the identified fork tip edge obtains two reference points for indicating the position of the fork tip; a connection of two reference points indicating the position of the fork point; the fork tip is monitored for climbing through the connecting line parallel to the X axis.
3. The method of video-based switch point crawling monitoring as in claim 1, wherein said switch point edge profile identification of step three further comprises the steps of:
SS1: manually attaching a plurality of label points on one side of the switch rail;
SS2: photographing the tag point by a high-definition camera at one side of the switch rail; and marks it as an initial tag point, denoted by APi, i=1, 2, … …, n;
SS3: when the switch point is monitored, the tag point is photographed again and marked as a real-time tag point, and the real-time tag point is denoted by BPi;
SS4: matching the real-time tag points with the initial tag points one by one; when it is deviated, it means that the point rail climbs longitudinally in the direction of the stock rail.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911258437.4A CN111144235B (en) | 2019-12-10 | 2019-12-10 | Switch tongue track crawling monitoring method based on video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911258437.4A CN111144235B (en) | 2019-12-10 | 2019-12-10 | Switch tongue track crawling monitoring method based on video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111144235A CN111144235A (en) | 2020-05-12 |
| CN111144235B true CN111144235B (en) | 2023-06-02 |
Family
ID=70518060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911258437.4A Active CN111144235B (en) | 2019-12-10 | 2019-12-10 | Switch tongue track crawling monitoring method based on video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111144235B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114312905B (en) * | 2021-11-25 | 2023-01-13 | 广州智为科技发展有限公司 | Switch point rail form image real-time supervision device |
| CN116279650B (en) * | 2023-05-25 | 2023-08-18 | 中铁四局集团有限公司 | Switch tongue close contact detection method and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0710002A (en) * | 1993-06-29 | 1995-01-13 | Railway Technical Res Inst | Turnout tip locking device and drive method |
| EP2165915A2 (en) * | 2008-09-23 | 2010-03-24 | VolkerRail Nederland BV | Monitoring a turnout of a railway or tramway line |
| CN104109991A (en) * | 2014-06-23 | 2014-10-22 | 杭州慧景科技有限公司 | Turnout rail creeping amount monitoring device |
| CN203964858U (en) * | 2014-06-23 | 2014-11-26 | 杭州慧景科技有限公司 | The closely connected monitoring device of a kind of railroad turnout steel rail |
| CN104236498A (en) * | 2014-09-30 | 2014-12-24 | 中铁山桥集团有限公司 | Method for measuring end size of switch rail |
| CN105472355A (en) * | 2016-01-13 | 2016-04-06 | 江苏金晓电子信息股份有限公司 | Switch machine notch monitoring system and method based on binocular visual processing identification |
| DE102014220778A1 (en) * | 2014-10-14 | 2016-04-14 | Robert Bosch Gmbh | Method for monitoring track sections in a rail vehicle |
| CN107476151A (en) * | 2017-08-21 | 2017-12-15 | 四川西南交大铁路发展股份有限公司 | Point tongue based on image recognition is creeped quantity monitoring method, device and its system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8345948B2 (en) * | 2009-09-11 | 2013-01-01 | Harsco Corporation | Automated turnout inspection |
-
2019
- 2019-12-10 CN CN201911258437.4A patent/CN111144235B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0710002A (en) * | 1993-06-29 | 1995-01-13 | Railway Technical Res Inst | Turnout tip locking device and drive method |
| EP2165915A2 (en) * | 2008-09-23 | 2010-03-24 | VolkerRail Nederland BV | Monitoring a turnout of a railway or tramway line |
| CN104109991A (en) * | 2014-06-23 | 2014-10-22 | 杭州慧景科技有限公司 | Turnout rail creeping amount monitoring device |
| CN203964858U (en) * | 2014-06-23 | 2014-11-26 | 杭州慧景科技有限公司 | The closely connected monitoring device of a kind of railroad turnout steel rail |
| CN104236498A (en) * | 2014-09-30 | 2014-12-24 | 中铁山桥集团有限公司 | Method for measuring end size of switch rail |
| DE102014220778A1 (en) * | 2014-10-14 | 2016-04-14 | Robert Bosch Gmbh | Method for monitoring track sections in a rail vehicle |
| CN105472355A (en) * | 2016-01-13 | 2016-04-06 | 江苏金晓电子信息股份有限公司 | Switch machine notch monitoring system and method based on binocular visual processing identification |
| CN107476151A (en) * | 2017-08-21 | 2017-12-15 | 四川西南交大铁路发展股份有限公司 | Point tongue based on image recognition is creeped quantity monitoring method, device and its system |
Non-Patent Citations (2)
| Title |
|---|
| 张宇宁 ; 谢琦 ; .一种基于机器视觉的铁路道岔检测方法.计算机应用与软件.2015,(01),全文. * |
| 郑霄 ; 王俊峰 ; .基于图像处理技术的转辙机实时在线监测研究.铁道通信信号.2009,(06),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111144235A (en) | 2020-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110254468B (en) | Intelligent online detection device and detection method for track surface defects | |
| CN111144235B (en) | Switch tongue track crawling monitoring method based on video | |
| CN110555361B (en) | Image processing method for lane classification | |
| CN103324198B (en) | A kind of truck positioning automated induction systems based on computer vision technique and using method thereof | |
| CN109300125B (en) | Method for extracting pantograph slide plate image of urban rail train in complex environment | |
| CN117358615B (en) | Automatic code-spraying printing defect detection method and system | |
| CN104809449B (en) | Track dotted line line of demarcation automatic testing method suitable for highway video monitoring system | |
| CN105472355A (en) | Switch machine notch monitoring system and method based on binocular visual processing identification | |
| CN107703513A (en) | A kind of novel non-contact contact net relative position detection method based on image procossing | |
| CN110567680A (en) | Track fastener looseness detection method based on angle comparison | |
| CN105023018A (en) | Jet code detection method and system | |
| CN110084171B (en) | Detection device and detection method for foreign matters on top of subway train | |
| CN110490163B (en) | Intelligent processing method and device for railway video data | |
| CN113256723B (en) | Automatic detection method for pantograph lifting time and pantograph head displacement curve | |
| WO2022170633A1 (en) | Rail transit vehicle collision avoidance detection method based on vision and laser ranging | |
| Wang et al. | An inverse projective mapping-based approach for robust rail track extraction | |
| CN117570880A (en) | Profile straightness detection method | |
| CN114830180A (en) | Automatic checking system for camera | |
| CN112037207A (en) | Method for detecting closing fault of automatic brake valve plug handle during railway wagon derailment | |
| CN207923179U (en) | Intelligent patrol detection vehicle | |
| CN110634122A (en) | Fastener elastic strip displacement detection method | |
| CN116152147A (en) | Connector pin position deviation detection method | |
| CN110329913B (en) | Device, method and system for controlling lifting equipment for positioning cart | |
| CN114666550A (en) | Contact point parameter detection device and method of close contact checker based on image processing | |
| CN115063448A (en) | Multi-target tracking method for oilfield operation scene |
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 |