CN102628814B - Automatic detection method of steel rail light band abnormity based on digital image processing - Google Patents
Automatic detection method of steel rail light band abnormity based on digital image processing Download PDFInfo
- Publication number
- CN102628814B CN102628814B CN 201210047032 CN201210047032A CN102628814B CN 102628814 B CN102628814 B CN 102628814B CN 201210047032 CN201210047032 CN 201210047032 CN 201210047032 A CN201210047032 A CN 201210047032A CN 102628814 B CN102628814 B CN 102628814B
- Authority
- CN
- China
- Prior art keywords
- rail
- light belt
- zone
- image
- straight line
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses an automatic detection method of steel rail light band abnormity based on digital image processing, comprising the following steps of: analyzing track images taken with a camera by the adoption of a digital image processing technique and detecting whether a steel rail light band is abnormal or not; preliminarily positioning a steel rail area in the images by the utilization of methods of image enhancement, edge detection, line detection and the like; extracting the light band area on the top of the steel rail through methods of pattern recognition, image segmentation, threshold processing and the like; and finally carrying out statistics and analysis on the extracted steel rail light band area and identifying whether it becomes abnormal or not. The method provided by the invention can be used to efficiently, automatically and intelligently detect stability of the steel rail, effectively minimize manpower input, reduce detection time, guarantee detection accuracy and effectively guarantee safety during high-speed operation of trains.
Description
Technical field
The present invention relates to the track stationary performance and measure, especially the collection rail image of rail in high speed railway stationarity monitoring detect the whether method of abnormal of rail light belt wherein.
Background technology
Along with the development of high-speed railway, train running speed is more and more faster, and the requirement of track stationarity is improved constantly.Yet, being subject to the impact of the factors such as building railway, geographical environment and train operation, the stationarity of track there will be problem unavoidably.Because the stationarity of track is understood the security of operation that directly affect train, therefore, for the detection of track stationarity, concerning the security of the lives and property of numerous people.
The general detection to the track stationarity is that the light belt by staying in train travelling process on rail is judged.The rail light belt refer to when train wheel at rail surface, roll, while sliding, wheel rim and rail interphase interaction, the bright trace stayed on rail.It is that mode by manual observation is judged that traditional rail light belt detects., stop transport at night by day always in running status due to bullet train, therefore can only, by night traffic less situation, use detecting lamp to detect one by one the rail artificially.Whole nation high speed railway track will reach 1.3 ten thousand kilometers in 2012, be at short notice method by manual detection judge whether abnormal difficulty very of rail light belt.Simultaneously, due to the people at night in fatigue state, add the reasons such as glazed thread, easily the detection of track stationarity is produced to the undetected and flase drop.This makes the safety of train when high-speed cruising be difficult to obtain effective guarantee.A kind of efficient, automatic, intelligent rail stationarity detection method necessitates.
Whether in the process of manual detection, whether normally observe the rail surface light belt has the width of obvious ripple, light belt increase or reduce to judge that whether track is steady.Rail surface light belt feature is taken into full account, in conjunction with computer vision, digital image processing techniques, can effectively reduce human input, reduce detection time, and ensure the accuracy rate detected.
Summary of the invention
Above deficiency in view of prior art, the present invention aims to provide a kind of computer vision, digital image processing techniques utilized and analyzes the track picture that camera is taken, and detect the whether method of abnormal of its rail light belt, make it to overcome the above deficiency of prior art, complete efficiently, automatically, intelligently Detection task.
Purpose of the present invention realizes by following means.
1) a kind of abnormal automatic testing method of rail light belt based on Digital Image Processing, adopt digital image processing techniques to analyze the track picture that camera takes and detect whether abnormal of its rail light belt, its pack processing contains following steps: by actual measurement is carried out in a large amount of normal steel track surfaces zone, measure the width in the non-light belt of width and the both sides zone in light belt zone with rule, obtain the width threshold value (ratio that width threshold value is light belt zone and non-light belt zone) in light belt zone according to the mass data of statistics, then calculate the undulating quantity of every section rail area light belt edge according to variance, obtain the ripple threshold value in light belt zone,
2) read the rail picture of shooting, use the method for rim detection, extract the Edge texture in the rail image;
3) edge image step 2 obtained carries out filtering, reduces the noise in image;
4) if the roughly direction of known rail in the middle of image, by rotation, make rail approach in image vertical, directly go to step 6, otherwise, the longest straight line in the Edge texture image that the method for use straight-line detection finds step 3 to obtain, the direction by its direction as rail in image;
5) if the rail direction state close to the vertical shape that step 4 obtains directly goes to step 6, otherwise the Edge texture image that step 3 is obtained is rotated, and makes the rail direction approach vertical;
6) from left to right extract in the Edge texture image and approach the straight line on vertical direction, merge too approaching in vertical direction parallel lines, then the length by straight line with and at textural characteristics, the color characteristic of former figure corresponding region, judge whether this straight line is the edge line of rail surface both sides, if, preserve the straight line parameter, final, the edge line of acquisition rail surface both sides.
7) extract the straight line that approaches horizontal direction in the Edge texture image, choose wherein two and non-neighbours' the longest straight line, as in image perpendicular to the reference line of track;
8) the rail surface both sides of the edge straight line extracted and the straight line that approaches horizontal direction are intersected in twos, obtain four intersection points;
9) four quadrilaterals corresponding to intersection point of step 8 gained are mapped as to a rectangle in new picture, four summits of this rectangle are corresponding one by one with four intersection points of step 8 gained respectively;
10) utilize the corresponding relation on four intersection points of step 8 gained and step 9 rectangle four summits, calculate the transition matrix that quadrilateral is converted to rectangle;
11) utilize step 10 gained transition matrix that rail image mapped to step 9 gained is newly schemed, between two vertical direction limits of new figure rectangle, zone is the Rail Surface zone;
12) utilize the Color Distribution Features in rail light belt and non-light belt zone, obtain light belt zone and the non-light belt zone in rail zone in conjunction with threshold process, extract Rail Surface light belt zone and the regional edge line had a common boundary of non-light belt, utilize the ripple Threshold Analysis light belt of statistics in advance whether to have ripple, if there is ripple, go to step 14;
13) calculate the light belt zone in the shared width ratio of Rail Surface, utilize the width threshold value of adding up in advance to analyze the light belt zone whether wide or narrow, if the light belt peak width is normal, prompting is normal, if the light belt peak width is abnormal, go to step 14, go to step 2;
The physical location information of 14) track record surface abnormalities, and recording exceptional point, send the abnormal prompt signal.
Adopt method of the present invention, utilize rail zone in the method Primary Location images such as figure image intensifying, rim detection, straight-line detection; Then by pattern-recognition, image cut apart, the method such as threshold process extracts rail surface light belt zone; Finally, its whether abnormal is added up, is analyzed and identified in extracted rail light belt zone.Detection rail stationarity that can be efficient, automatic, intelligent, effectively reduce human input, reduce detection time, and ensure the accuracy rate detected, and makes the safety of train when high-speed cruising obtain effective guarantee.
Accompanying drawing is described as follows:
Fig. 1 is the light belt of the required Four types detected in rail detects.In figure, a means that the rail surface light belt is normal, and b means that the rail surface light belt reduces, and c means that the rail surface light belt increases, and d means that the rail surface light belt has ripple.The reason that produces b is train at high-speed cruising to " de-luxe compartment " while locating, car body produces acceleration upwards, wheel suspension off-load, and on rail, dynamic vertical power reduces, the contact area of wheel and rail surface reduces, and causes the light belt that wheel stays on rail surface to reduce.The reason that produces c is that train high-speed cruising to " low-lying " located, and on rail, dynamic vertical power increases, and wheel and rail surface contact area increase, and the light belt that causes wheel to stay on rail surface increases.The reason that produces d is that loosening or track irregularity occur track, makes train roll in operational process, causes the light belt that wheel stays on rail surface that ripple is arranged.
Fig. 2 is that the camera that the present invention designs tilts to take the schematic diagram of track picture in track edge orbital direction.
The schematic diagram of Fig. 3 to be the camera that designs of the present invention take directly over perpendicular to track track picture.
Fig. 4 be the present invention design at rail, certain inclination and deformation are arranged in picture the time, calculate the algorithm flow chart of the transition matrix for proofreading and correct the rail picture.
Fig. 5 is that the present invention utilizes former figure and transition matrix to proofread and correct the rail picture, then extracts and identify the whether algorithm flow chart of abnormal of light belt zone in rail.
Embodiment
Below in conjunction with accompanying drawing and concrete embodiment, the present invention is described in further detail.
The concrete steps of this method are as follows:
The first step, the measurement of rail surface light belt width fluctuation range.
1) by actual measurement is carried out in a large amount of normal steel track surfaces zone, measure the width in the non-light belt of width and the both sides zone in light belt zone with rule, obtain the width threshold value in light belt zone according to the mass data of statistics, then calculate the undulating quantity at light belt edge in every section rail zone according to variance, obtain the ripple threshold value in light belt zone;
Second step, gather the rail image, extracts rail surface both sides of the edge line and, perpendicular to the reference of rail, calculate transition matrix.
2) collection of rail image suggestion, as shown in Figures 2 and 3, Fig. 2 takes the image of rail along the rail direction in the rail both sides, Fig. 3 is photographs rail image in the vertical direction, if it is better that camera frame is located to track checking car photographs effect, can use the multiple Boundary extracting algorithms such as Canny, Sobel, Prewitt, Robert, small echo, Qu Bo, profile ripple to the rail image gathered, obtain the edge image of rail image;
3) rail edge image step 2 obtained, need to remove noise in edge image, image can be carried out to smoothing processing, connective process etc., gets rid of noise spot and make at the nearer but discontinuous straight line of edge extracting middle distance and interconnect;
4) if the roughly direction of known rail in the middle of image makes rail approach by rotation vertical, directly go to step 6 in image, otherwise, can use Hough straight-line detection, Radon conversion straight-line detection etc., extract longer straight line in the rail image;
5) step 4 is chosen to the longest straight line and be the rail direction, if rail direction state close to the vertical shape directly goes to step 6, otherwise the rail image that step 4 is obtained is rotated, and makes the rail direction approach vertical;
6) from left to right extract in the Edge texture image and approach the straight line on vertical direction, merge too approaching in vertical direction parallel lines, then the length by straight line with and at textural characteristics, the color characteristic of former figure corresponding region, judge whether this straight line is the edge line of Rail Surface both sides, if, preserve the straight line parameter, final, the edge line of acquisition Rail Surface both sides;
7) extract the straight line that approaches horizontal direction in the Edge texture image, choose wherein two and non-neighbours' the longest straight line, as in image perpendicular to the reference line of track;
8) calculate the straight line of rail both sides and the intersection point of horizontal linear, obtain needing the tetragonal apex coordinate of conversion, make it, the upper left point is designated as (x1, y1), the lower-left point coordinate is (x2, y2), and the lower-right most point coordinate is (x3, y3), the upper right point coordinate is (x4, y4);
9) four summits that obtain according to step 8 build new rectangle, make its upper left corner coordinate for (x2, y1), and lower left corner coordinate is (x2, y2), and lower right corner coordinate is (x3, y2), and upper right corner coordinate is (x3, y1);
10) use perspective transform calculation procedure 8 gained quadrilaterals to be converted to the transition matrix of the new rectangle of step 9 gained;
The 3rd step, to the rail correct image, the rail zone after extract proofreading and correct, then extract rail light belt zone, judges whether abnormal of rail light belt.
11) utilize rail image mapped to step 9 gained after step 10 gained transform matrix calculations rail is corrected newly to scheme, between new two vertical direction line segments of figure rectangle, zone is the Rail Surface zone;
12) utilize the Color Distribution Features in rail light belt and non-light belt zone, obtain light belt zone and the non-light belt zone in rail zone in conjunction with threshold process, extract Rail Surface light belt zone and the regional edge line had a common boundary of non-light belt, utilize the ripple Threshold Analysis light belt that step 1 is added up in advance whether to have ripple, if there is ripple, go to step 14;
13) calculate the light belt zone in the shared width ratio of Rail Surface, utilize the width threshold value of adding up in advance to analyze the light belt zone whether wide or narrow, if the light belt peak width is normal, prompting is normal, if the light belt peak width is abnormal, go to step 14, go to step 2;
The physical location information of 14) track record surface abnormalities, and recording exceptional point, send abnormal prompt information.
Embodiment
Below case step explanation of the present invention:
1) by manual type, actual measurement is carried out in a large amount of normal steel track surfaces zone, measure the width in the non-light belt of width and the both sides zone in light belt zone with rule, obtain the width threshold value in light belt zone (according to actual measurement according to the mass data of statistics, obtaining width threshold value is between 0.5 to 0.6), then calculate the undulating quantity at light belt edge in every section rail zone according to variance, obtain the ripple threshold value (according to actual measurement, obtain the ripple threshold value and be less than 0.01) in light belt zone;
2) collection of rail image as shown in Figure 2, is used the Canny rim detection to obtain the edge image of rail image;
3) rail edge image step 2 obtained, used Smooth smoothing processing (image being carried out to core size is the Gaussian convolution of 3*3) eliminating noise spot to make at the nearer but discontinuous straight line of edge extracting middle distance and interconnect;
4) if the roughly direction of known rail in the middle of image makes rail approach by rotation vertical, directly go to step 6 in image, otherwise, can use the Hough straight-line detection to extract longer straight line in the rail image;
5) step 4 is chosen to the longest straight line and be the rail direction, if rail direction state close to the vertical shape directly goes to step 6, otherwise the rail image that step 4 is obtained is rotated, and makes the rail direction approach vertical;
6) from left to right extract in the Edge texture image and approach the straight line on vertical direction, merge too approaching in vertical direction parallel lines, then the length by straight line with and at the color characteristic of former figure corresponding region, judge whether this straight line is the edge line of Rail Surface both sides, if, preserve the straight line parameter, final, the edge line of acquisition Rail Surface both sides;
7) extract the straight line that approaches horizontal direction in the Edge texture image, choose wherein two and non-neighbours' the longest straight line, as in image perpendicular to the reference line of track;
8) calculate the straight line of rail both sides and the intersection point of horizontal linear, obtain needing the tetragonal apex coordinate of conversion, make it, the upper left point is designated as (x1, y1), the lower-left point coordinate is (x2, y2), and the lower-right most point coordinate is (x3, y3), the upper right point coordinate is (x4, y4);
9) four summits that obtain according to step 8 build new rectangle, make its upper left corner coordinate for (x2, y1), and lower left corner coordinate is (x2, y2), and lower right corner coordinate is (x3, y2), and upper right corner coordinate is (x3, y1);
10) use perspective transform calculation procedure 8 gained quadrilaterals to be converted to the transition matrix of the new rectangle of step 9 gained;
11) utilize rail image mapped to step 9 gained after step 10 gained transform matrix calculations rail is corrected newly to scheme, between new two vertical direction line segments of figure rectangle, zone is the Rail Surface zone;
12) utilize rail image mapped to step 9 gained after step 10 gained transform matrix calculations rail is corrected newly to scheme, between new two vertical direction line segments of figure rectangle, zone is the Rail Surface zone;
13) calculate the light belt zone in the shared width ratio of Rail Surface, utilize the width threshold value of adding up in advance to analyze the light belt zone whether wide or narrow, if the light belt peak width is normal, prompting is normal, if the light belt peak width is abnormal, go to step 14, go to step 2;
Track record surface abnormalities, and the physical location information of recording exceptional point, send abnormal prompt information.
Claims (1)
1. the abnormal automatic testing method of rail light belt based on Digital Image Processing, adopt digital image processing techniques to analyze the track picture that camera takes and detect whether abnormal of its rail light belt, and its pack processing is containing following steps:
1) by actual measurement is carried out in a large amount of normal steel track surfaces zone, measure the width in the non-light belt of width and the both sides zone in light belt zone with rule, obtain the width threshold value in light belt zone according to the mass data of statistics, the ratio that described width threshold value is light belt zone and non-light belt zone; Then calculate the undulating quantity of every section rail area light belt edge according to variance, obtain the ripple threshold value in light belt zone;
2) read the rail picture of shooting, use the method for rim detection, extract the Edge texture in the rail image;
3) edge image step 2 obtained carries out filtering, reduces the noise in image;
4) if the roughly direction of known rail in the middle of image makes rail approach by rotation vertical, directly go to step 6 in image; Otherwise, the longest straight line in the Edge texture image that the method for use straight-line detection finds step 3 to obtain, the direction by its direction as rail in image;
5) if the rail direction state close to the vertical shape that step 4 obtains directly goes to step 6; Otherwise the Edge texture image that step 3 is obtained is rotated, make the rail direction approach vertical;
6) from left to right extract in the Edge texture image and approach the straight line on vertical direction, merge too approaching in vertical direction parallel lines, then the length by straight line with and at textural characteristics, the color characteristic of former figure corresponding region, judge whether this straight line is the edge line of rail surface both sides, if preserve the straight line parameter; Finally, obtain the edge line of rail surface both sides;
7) extract the straight line that approaches horizontal direction in the Edge texture image, choose wherein two and non-neighbours' the longest straight line, as in image perpendicular to the reference line of track;
8) the rail surface both sides of the edge straight line extracted and the straight line that approaches horizontal direction are intersected in twos, obtain four intersection points;
9) four quadrilaterals corresponding to intersection point of step 8 gained are mapped as to a rectangle in new picture, four summits of this rectangle are corresponding one by one with four intersection points of step 8 gained respectively;
10) utilize the corresponding relation on four intersection points of step 8 gained and step 9 rectangle four summits, calculate the transition matrix that quadrilateral is converted to rectangle;
11) utilize step 10 gained transition matrix that rail image mapped to step 9 gained is newly schemed, between two vertical direction limits of new figure rectangle, zone is the Rail Surface zone;
12) utilize the Color Distribution Features in rail light belt and non-light belt zone, obtain light belt zone and the non-light belt zone in rail zone in conjunction with threshold process, extract Rail Surface light belt zone and the regional edge line had a common boundary of non-light belt, utilize the ripple Threshold Analysis light belt of statistics in advance whether to have ripple, if there is ripple, go to step 14;
13) calculate the light belt zone in the shared width ratio of Rail Surface, utilize the width threshold value of adding up in advance to analyze the light belt zone whether wide or narrow, if the light belt peak width is normal, prompting is normal, if the light belt peak width is abnormal, go to step 14, go to step 2;
The physical location information of 14) track record surface abnormalities, and recording exceptional point, send the abnormal prompt signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201210047032 CN102628814B (en) | 2012-02-28 | 2012-02-28 | Automatic detection method of steel rail light band abnormity based on digital image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201210047032 CN102628814B (en) | 2012-02-28 | 2012-02-28 | Automatic detection method of steel rail light band abnormity based on digital image processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102628814A CN102628814A (en) | 2012-08-08 |
| CN102628814B true CN102628814B (en) | 2013-12-18 |
Family
ID=46587113
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201210047032 Expired - Fee Related CN102628814B (en) | 2012-02-28 | 2012-02-28 | Automatic detection method of steel rail light band abnormity based on digital image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102628814B (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102013109915B4 (en) * | 2013-09-10 | 2015-04-02 | Thyssenkrupp Steel Europe Ag | Method and device for checking an inspection system for detecting surface defects |
| CN104197842A (en) * | 2014-09-15 | 2014-12-10 | 常州宝仪机电设备有限公司 | Image type plate width measurement and control system |
| CN104713486B (en) * | 2015-03-20 | 2017-08-15 | 西安交通大学 | One kind is puted the palms together before one product width on-line measuring device and method |
| CN104792270A (en) * | 2015-04-16 | 2015-07-22 | 石家庄铁路职业技术学院 | Deep foundation pit detection method based on white-light digital image frequency domain analysis method |
| CN105466941B (en) * | 2015-11-27 | 2018-08-14 | 中国铁道科学研究院 | A kind of steel rail detection method and equipment of the light belt image based on rail |
| CN106355185B (en) * | 2016-08-30 | 2019-07-16 | 兰州交通大学 | A kind of method in rapidly extracting Rail Surface region under vibration condition |
| CN106428558B (en) * | 2016-11-28 | 2018-06-01 | 北京交通大学 | A kind of track synthesis method for inspecting based on the dual-purpose unmanned plane of sky-rail |
| CN106494611B (en) * | 2016-11-28 | 2018-07-13 | 北京交通大学 | A kind of dual-purpose patrol unmanned machine of sky rail |
| CN107194921B (en) * | 2017-05-19 | 2020-04-07 | 华夏高铁技术有限公司 | Automatic identification method and device for rail surface light band of steel rail |
| CN107238605B (en) * | 2017-06-08 | 2021-07-02 | 广东嘉铭智能科技有限公司 | Method and device for detecting defects on outer surface of cylindrical workpiece |
| CN107264570B (en) * | 2017-07-25 | 2019-07-05 | 西南交通大学 | Steel rail light band distribution detecting device and method |
| CN108510475B (en) * | 2018-03-09 | 2022-03-29 | 南京合迈美家智能科技有限公司 | Method and system for measuring muscle tendon knots in muscle continuous ultrasonic images |
| CN109117757B (en) * | 2018-07-27 | 2022-02-22 | 四川大学 | Method for extracting guy cable in aerial image |
| CN111105409B (en) * | 2019-12-27 | 2023-08-08 | 中国铁道科学研究院集团有限公司 | Automatic detection method, equipment and system for steel rail light bands based on image processing |
| CN113112474B (en) * | 2021-04-12 | 2022-11-18 | 东南大学 | Rail positioning method for turnout and non-turnout track images |
| CN114384078B (en) * | 2021-12-31 | 2023-11-24 | 武汉大学 | Method and equipment for detecting wave abrasion damage of steel rail in positioning mode |
| CN114166855B (en) * | 2022-02-11 | 2022-04-29 | 南京邮电大学 | Real-time rail defect detection method |
| CN116109638B (en) * | 2023-04-13 | 2023-07-04 | 中铁四局集团有限公司 | Rail break detection method and system |
| CN117253066B (en) * | 2023-11-20 | 2024-02-27 | 西南交通大学 | Rail surface state identification method, device, equipment and readable storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5621645A (en) * | 1995-01-24 | 1997-04-15 | Minnesota Mining And Manufacturing Company | Automated lane definition for machine vision traffic detector |
| DE19636028C1 (en) * | 1996-09-05 | 1997-11-20 | Daimler Benz Ag | Stereo image object detection, esp. for road vehicles |
| CN1645037A (en) * | 2004-12-31 | 2005-07-27 | 天津大学 | Partitioned pointing method for three-dimensional scanning measurement system by light band method |
| WO2008091565A1 (en) * | 2007-01-23 | 2008-07-31 | Valeo Schalter & Sensoren Gmbh | Method and system for universal lane boundary detection |
| CN101620732A (en) * | 2009-07-17 | 2010-01-06 | 南京航空航天大学 | Visual detection method of road driving line |
-
2012
- 2012-02-28 CN CN 201210047032 patent/CN102628814B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN102628814A (en) | 2012-08-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102628814B (en) | Automatic detection method of steel rail light band abnormity based on digital image processing | |
| Liu et al. | A review of applications of visual inspection technology based on image processing in the railway industry | |
| CN106652445B (en) | A kind of road traffic accident method of discrimination and device | |
| CN104504388B (en) | A kind of pavement crack identification and feature extraction algorithm and system | |
| CN103123722B (en) | Road object detection method and system | |
| CN107025432B (en) | A kind of efficient lane detection tracking and system | |
| CN102508110B (en) | Texture-based insulator fault diagnostic method | |
| KR102017870B1 (en) | Real-time line defect detection system | |
| CN102999749B (en) | Based on the securing band violation event intelligent detecting method of Face datection | |
| CN110136447A (en) | Lane change of driving a vehicle detects and method for distinguishing is known in illegal lane change | |
| CN103034850B (en) | Trouble of moving freight car detection system (TFDS) block key loss fault automatic identification method | |
| CN111260629A (en) | Pantograph structure abnormity detection algorithm based on image processing | |
| CN105260713A (en) | Method and device for detecting lane line | |
| CN104608799A (en) | Information fusion technology based train wheel set tread damage online detection and recognition method | |
| CN103593649A (en) | Lane line detection method for lane departure early warning | |
| CN105426894B (en) | Railway plug pin image detecting method and device | |
| CN103295420A (en) | Method for recognizing lane line | |
| CN110862033A (en) | Intelligent early warning detection method applied to coal mine inclined shaft winch | |
| CN107179479B (en) | Visible light image-based transmission conductor strand scattering defect detection method | |
| CN105133471A (en) | Linear structured light pavement surface detection system-based pavement depth image production method | |
| TWI611961B (en) | Device, method, and computer-readable medium for analyzing lane line image | |
| Liu et al. | An efficient method for high-speed railway dropper fault detection based on depthwise separable convolution | |
| CN105718916A (en) | Lane line detection method based on Hough transform | |
| CN105241679A (en) | Method for detecting hidden fault of motor train unit | |
| CN111561967A (en) | Real-time online detection method and system for pantograph-catenary operation state |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20131218 Termination date: 20180228 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |