CN106156780A - The method getting rid of wrong report on track in foreign body intrusion identification - Google Patents
The method getting rid of wrong report on track in foreign body intrusion identification Download PDFInfo
- Publication number
- CN106156780A CN106156780A CN201610497533.4A CN201610497533A CN106156780A CN 106156780 A CN106156780 A CN 106156780A CN 201610497533 A CN201610497533 A CN 201610497533A CN 106156780 A CN106156780 A CN 106156780A
- Authority
- CN
- China
- Prior art keywords
- image
- foreign body
- track
- wrong report
- body intrusion
- 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
Classifications
-
- 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
- G06V10/443—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 by matching or filtering
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of method getting rid of wrong report on track in foreign body intrusion identification, comprise the steps: (1), quickly obtain video image；(2), the video image collected is carried out preliminary treatment；(3) two orbital paths, are obtained；(4) Dynamic Envelope of train, is drawn；(5), start training mode, in determining the environment not having foreign body intrusion, the equipment being defined as on track by the foreign body recognized, extract characteristics of image and be stored in wrong report feature database；(6), under vehicle normal operation mode, in Dynamic Envelope, take image block according to orbital path from the near to the remote carry out Characteristic Contrast, be confirmed whether it is foreign body intrusion.Have the beneficial effect that the method for the present invention can reduce the misinformation probability in track foreign body intrusion identification, get rid of in the middle of track or the equipment of both sides, Sign Board etc. are erroneously interpreted as the probability of foreign body intrusion, thus promoting the accuracy of identification, it is ensured that train runs the most in an orderly manner.
Description
Technical field
The present invention relates to foreign body intrusion on a kind of track and know method for distinguishing, be specifically related to foreign body intrusion identification on a kind of track
The middle method getting rid of wrong report；Belong to technical field of rail traffic.
Background technology
Along with the development of traffic new technique, detection foreign body intrusion and prediction danger are the indispensable important merits of intelligent vehicle
One of can.Front vehicles is carried out real-time automatic detection and identification for keeping safe distance between vehicles, preventing collision accident and have
Highly important meaning, is also the precondition of safety traffic.In traditional driving technology, in driver vehicle processes
Required information, substantially from vision, is inspired by this, has developed computer vision (also referred to as machine based on image procossing and pattern recognition
Device vision), it is possible to utilize image and image sequence to identify and cognitive three-dimensional world, make computer generation replace and surmount human vision
Some function of system.Step by step, acquisition of information means during machine vision has become current intelligent vehicle and safety assistant driving
Main path.
In the road foreign body intrusion of prior art detects, Cleaning Principle is to train based on characteristics of image and cascade classifier
Classify two kinds of algorithms, used three kinds of methods in car conour feature: one is that the Threshold segmentation of gray space carries out bottom of car
Shadow recognition, two is to utilize symmetry to come whether authentication image region is automobile, and three is to utilize HIS color space Threshold segmentation
Carry out the identification of automobile tail light.During cascade classifier training, need substantial amounts of sample, obtained by some by great amount of samples training
The cascade classifier of individual Weak Classifier composition, is converted to integrogram by inspection image during detection, then utilizes the classification of different proportion
Device window scanning entire image, obtains comprising the different size of image-region of automobile, finally by overlapping region merging technique, obtains
Final testing result.Owing to entire image being taken multiple scan, necessarily cause that image information to be processed is big, operation efficiency
Low, but the foreign body intrusion that above-mentioned traditional method detects is single, predominantly detects other vehicles on road, it is impossible to detect other
Foreign body intrusion, the non-foreign body intrusion object being similar to automobile cannot be got rid of, more cannot get rid of the facilities such as normal traffic mark,
It is easily caused erroneous judgement.
Summary of the invention
For solving the deficiencies in the prior art, it is an object of the invention to provide and on a kind of track, foreign body intrusion identification is got rid of
The method of wrong report, thus improve the accuracy to the foreign body intrusion identification occurred on track.
In order to realize above-mentioned target, the present invention adopts the following technical scheme that:
The method getting rid of wrong report on track in foreign body intrusion identification, comprises the steps:
(1), quickly obtained video image by video acquisition module and transmit to image pre-processing module；
(2), preliminary treatment is carried out by the image pre-processing module video image to collecting；
(3), the image through preliminary treatment is carried out between Hough transformation, binocular ranging, two rails by track detecting module successively
Familiar route judgement, rim detection, curve matching and smoothing processing obtain two orbital paths；
(4), Dynamic Envelope drafting module draws the dynamic of train from the near to the remote according to the pixel distance ratio between two rails
State envelope；
(5), start training mode, in determining the environment not having foreign body intrusion, the foreign body intrusion recognized is defined as
Equipment on track, extracts characteristics of image and is stored in wrong report feature database；
(6), under vehicle normal operation mode, in Dynamic Envelope, image block is taken from the near to the remote according to orbital path, will
Away from headstock level altitude the most nearby, a length of left rail to right gauge from orbital image block to be compared, be close to track to be compared
Image block front level altitude, a length of left rail to right gauge from orbital image block carry out Characteristic Contrast, feature is inconsistent
Preliminary identification orbital image block to be compared is foreign body intrusion outputting alarm；Again by special with the image in wrong report feature database for its feature
Levy and contrast one by one, mate with any feature in wrong report feature database, then get rid of the possibility of foreign body intrusion, do not mate, confirm as different
Thing invades limit.
Preferably, aforementioned video acquisition module includes: be divided into the left and right sides, synchronous acquisition in train head windshield
First photographic head of video image and second camera, be so able to ensure that and collect distance video image clearly.
More preferably, the preliminary treatment process of abovementioned steps (2) is specific as follows:
The RGB image received is converted to gray level image by image pre-processing module；
Scharr operator is used to be filtered operation, filter coefficient
Image smoothing operates, and image carries out Gaussian convolution, and core size is 3*3, standard deviation sigma=(n/2-1) * 0.3+
0.8, the wherein corresponding horizontal core of n or vertical core size；
Smooth rear image is carried out secondary scaling: use Gaussian pyramid decomposition to input picture down sample, head
First input picture Gaussian filter is carried out convolution, then by refusing the row and column down sample image of even number；Use
Input picture to up-sampling, is first passed through and inserts 0 even number line and even column in the picture, then by Gaussian pyramid decomposition
The image Gaussian filter obtained is carried out Gaussian convolution, and its median filter is multiplied by 4 and does interpolation, and output image is input picture
4 times of sizes, Gaussian convolution core size is 5*5；
Eliminate picture noise, split independent pictorial element, use kernel structure be rectangle, size be the structural elements of 3*3
Element, scans each pixel of bianry image, doing AND operation with the bianry image of structural element Yu covering, if being all 1, then tying
This pixel of composition picture is 1, is otherwise 0, makes bianry image reduce a circle, and this operation is in triplicate；Connect adjacent element, use
Kernel structure be rectangle, size be the structural element of 3*3, scan each pixel of bianry image, with structural element and covering
Bianry image does AND operation, if being all 0, then this pixel of structural images is 0, is otherwise 1, makes bianry image expand one
Circle, operates three times；
Image is carried out binarization operation,Wherein thresh is
100。
It is highly preferred that the detailed process of abovementioned steps (3) is: Hough transformation uses the built-in function of opencv
HoughLinesP, wherein rho is 1, theta be pi/180, threshold be 80；SGBM algorithm is used to calculate on left and right view
Parallax carries out three-dimensional reconstruction, then calculate two rectilineal intervals from；Conversion formula between camera coordinate system and world coordinate systemWherein (Xc, Yc, Zc) represents P point position under camera coordinate system, and (Xw, Yw, Zw) represents P point
Position under world coordinate system, R is spin matrix, and T is translation matrix, and the inside and outside parameter that R and T is demarcated by binocular camera obtains
Arrive；Calculate two included angle of straight line, tan θ ＜ 0.1；Rim detection extracts profile, fits to curve, finds the highest with straight line degree of overlapping
Curve, smoothed curve obtains left and right orbital path.
It is further preferred that the characteristics of image in step (5) is invariant moment features, comprise moment of the orign, central moment and normalizing
Combination after change, two dimensional image with f (x, y) represent, then its (p+q) rank moment of the orign is:Center
Square is:P, q=0,1,2 ..., x0,y0Centre coordinate for image；In order to
Make characteristics of image that translation, rotation and transformation of scale are had invariance, with zeroth order central moment, remaining each rank central moment is returned
One changes, and obtains image normalization central momentWherein,P+q=2,3... p, q=0,1,2,3 are taken.
Further, aforementioned not bending moment uses seven kinds of not bending moment result of calculations, and the definition of seven kinds of not bending moments is:
m1=η20+η02
m3=(η30-3η12)2+(η03-3η21)2
m4=(η30+η12)2+(η03+η21)2
m5=(η30-3η12)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η03-3η21)(η03+η21)
m6=(η20-η02)[(η30+η12)2-(η21+η03)2]+4η11(η30+η12)(η03+η21)
m7=(3 η21-η03)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η30-3η12)(η21+η03)[(η21+η03)2-3
(η30+η12)2]。
Yet further, the characteristic matching in abovementioned steps (6) is analog quantity tolerance,p
For image to be identified, q is image in wrong report feature database.
Further, wrong report feature database is stored in a hard disk, reports feature database under vehicle normal operation mode by mistake complete
Internal memory is read in portion, thus improves arithmetic speed.
The invention have benefit that: the wrong report that the method for the present invention can reduce in track foreign body intrusion identification is general
Rate, gets rid of in the middle of track or the equipment of both sides, Sign Board etc. are erroneously interpreted as the probability of foreign body intrusion, thus promotes foreign body and invade
The accuracy that limit identifies, it is ensured that train runs the most in an orderly manner.
Accompanying drawing explanation
Fig. 1 is the flow chart of the method getting rid of wrong report on the track of the present invention in foreign body intrusion identification.
Detailed description of the invention
Below in conjunction with the drawings and specific embodiments, the present invention made concrete introduction.
In describing the invention, it is to be understood that term " " center ", " longitudinally ", " laterally ", " on ", D score,
Orientation or the position relationship of the instruction such as "front", "rear", "left", "right", " vertically ", " level ", " top ", " end ", " interior ", " outward " are
Based on orientation shown in the drawings or position relationship, it is for only for ease of the description present invention and simplifies description rather than instruction or dark
The device or the element that show indication must have specific orientation, with specific azimuth configuration and operation, therefore it is not intended that right
The restriction of the present invention.Additionally, term " first ", " second " are only used for describing purpose, and it is not intended that instruction or hint relatively
Importance.
From the point of view of the flow chart of Fig. 1, the method that the track of the present embodiment is got rid of in foreign body intrusion identification wrong report, including
Following steps:
(1), by video acquisition module quickly obtain video image and transmit to image pre-processing module, wherein, video
Acquisition module includes being divided into the first photographic head and the second camera of the left and right sides in train head windshield, and the two is taken the photograph
As head is by the secondary development bag synchronous acquisition video image carried, and, photographic head should be installed on eminence, just can ensure that collection
To distance video image clearly.
(2), carrying out preliminary treatment by the image pre-processing module video image to collecting, detailed process is:
The RGB image received is converted to gray level image by image pre-processing module；
Scharr operator is used to be filtered operation, filter coefficient
Image smoothing operates, and image carries out Gaussian convolution, and core size is 3*3, standard deviation sigma=(n/2-1) * 0.3+
0.8, the wherein corresponding horizontal core of n or vertical core size；
Smooth rear image is carried out secondary scaling: use Gaussian pyramid decomposition to input picture down sample, head
First input picture Gaussian filter is carried out convolution, then by refusing the row and column down sample image of even number；Use
Input picture to up-sampling, is first passed through and inserts 0 even number line and even column in the picture, then by Gaussian pyramid decomposition
The image Gaussian filter obtained is carried out Gaussian convolution, and its median filter is multiplied by 4 and does interpolation, and output image is input picture
4 times of sizes, Gaussian convolution core size is 5*5；
Eliminate picture noise, split independent pictorial element, use kernel structure be rectangle, size be the structural elements of 3*3
Element, scans each pixel of bianry image, doing AND operation with the bianry image of structural element Yu covering, if being all 1, then tying
This pixel of composition picture is 1, is otherwise 0, makes bianry image reduce a circle, and this operation is in triplicate；Connect adjacent element, use
Kernel structure be rectangle, size be the structural element of 3*3, scan each pixel of bianry image, with structural element and covering
Bianry image does AND operation, if being all 0, then this pixel of structural images is 0, is otherwise 1, makes bianry image expand one
Circle, operates three times；
Image is carried out binarization operation,Wherein thresh is
100。
(3), the image through preliminary treatment is carried out between Hough transformation, binocular ranging, two rails by track detecting module successively
Familiar route judgement, rim detection, curve matching and smoothing processing obtain two orbital paths, and detailed process is:
Hough transformation uses the built-in function HoughLinesP of opencv, and wherein rho is 1, and theta is pi/180,
Threshold is 80；Left and right view uses SGBM algorithm calculate parallax and carries out three-dimensional reconstruction, then calculate two rectilineal intervals from；
Conversion formula between camera coordinate system and world coordinate systemWherein (Xc, Yc, Zc) represents that P point exists
Position under camera coordinate system, (Xw, Yw, Zw) represents P point position under world coordinate system, and R is spin matrix, and T is flat
Moving matrix, the inside and outside parameter that R and T is demarcated by binocular camera obtains；Calculate two included angle of straight line, tan θ ＜ 0.1；Rim detection is extracted
Profile, fits to curve, finds the curve the highest with straight line degree of overlapping, and smoothed curve obtains left and right orbital path.
(4), Dynamic Envelope drafting module draws the dynamic of train from the near to the remote according to the pixel distance ratio between two rails
State envelope；
(5), start training mode, in determining the environment not having foreign body intrusion, the foreign body intrusion recognized is defined as
Equipment on track, extracts characteristics of image and is stored in wrong report feature database, and wrong report feature database is stored in a hard disk with data mode, side
Call the most at any time；
In this step, the characteristics of image being stored in wrong report feature database is invariant moment features, comprises moment of the orign, central moment and returns
One change after combination, two dimensional image with f (x, y) represent, then its (p+q) rank moment of the orign is:In
Heart square is:P, q=0,1,2 ..., x0,y0Centre coordinate for image；
In order to make characteristics of image that translation, rotation and transformation of scale are had invariance, the present embodiment is used zeroth order central moment
Remaining each rank central moment is normalized, obtains image normalization central momentWherein,P+q=
2,3... p, q=0,1,2,3 are taken.
Bending moment does not uses seven kinds of not bending moment result of calculations, and the definition of seven kinds of not bending moments is:
m1=η20+η02
m3=(η30-3η12)2+(η03-3η21)2
m4=(η30+η12)2+(η03+η21)2
m5=(η30-3η12)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η03-3η21)(η03+η21)
m6=(η20-η02)[(η30+η12)2-(η21+η03)2]+4η11(η30+η12)(η03+η21)
m7=(3 η21-η03)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η30-3η12)(η21+η03)[(η21+η03)2-3
(η30+η12)2]。
By these seven kinds not bending moment result of calculation can extract characteristics of image exactly, improve the standard of subsequent characteristics comparison
Really property and reliability.
(6), under vehicle normal operation mode, wrong report feature database all reads in internal memory, thus improves arithmetic speed.According to rail
Path takes image block from the near to the remote in Dynamic Envelope, will be away from headstock level altitude the most nearby, a length of left rail to right rail
The orbital image block to be compared of distance, with next-door neighbour orbital image block front to be compared level altitude, a length of left rail to right gauge
From orbital image block carry out Characteristic Contrast, inconsistent preliminary of feature assert that orbital image block to be compared is foreign body intrusion defeated
Go out alarm；Again its feature is contrasted one by one with the characteristics of image in wrong report feature database, mates with any feature in wrong report feature database,
Then get rid of the possibility of foreign body intrusion, do not mate, confirm as foreign body intrusion；
In this step, characteristic matching is analog quantity tolerance,P is image to be identified, and q is
Image in wrong report feature database.
To sum up, the method for the present invention can reduce the misinformation probability in track foreign body intrusion identification, get rid of in the middle of track or
The equipment of both sides, Sign Board etc. are erroneously interpreted as the probability of foreign body intrusion, thus promote the accuracy of foreign body intrusion identification, really
Protect train to run the most in an orderly manner.
In describing the invention, it should be noted that unless otherwise clearly defined and limited, term " is installed ", " phase
Even ", " connection " should be interpreted broadly, for example, it may be fixing connection, it is also possible to be to removably connect, or be integrally connected；Can
To be mechanical connection, it is also possible to be electrical connection；Can be to be joined directly together, it is also possible to be indirectly connected to by intermediary, Ke Yishi
The connection of two element internals.For the ordinary skill in the art, can understand that above-mentioned term is at this with concrete condition
Concrete meaning in invention.
In the description of this specification, reference term " embodiment ", " some embodiments ", " example ", " specifically show
Example " or the description of " some examples " etc. means to combine this embodiment or example describes specific features, structure, material or spy
Point is contained at least one embodiment or the example of the present invention.In this manual, to the schematic representation of above-mentioned term not
Necessarily refer to identical embodiment or example.And, the specific features of description, structure, material or feature can be any
One or more embodiments or example in combine in an appropriate manner.
The ultimate principle of the present invention, principal character and advantage have more than been shown and described.The technical staff of the industry should
Understanding, above-described embodiment limits the present invention the most in any form, and the mode of all employing equivalents or equivalent transformation is obtained
Technical scheme, all falls within protection scope of the present invention.
Claims (8)
1. the method getting rid of wrong report on track in foreign body intrusion identification, it is characterised in that comprise the steps:
(1), quickly obtained video image by video acquisition module and transmit to image pre-processing module；
(2), preliminary treatment is carried out by the image pre-processing module video image to collecting；
(3), track detecting module carries out familiar route between Hough transformation, binocular ranging, two rails successively to the image through preliminary treatment
Judgement, rim detection, curve matching and smoothing processing obtain two orbital paths；
(4), Dynamic Envelope drafting module draws the dynamic bag of train from the near to the remote according to the pixel distance ratio between two rails
Winding thread；
(5), start training mode, in determining the environment not having foreign body intrusion, the foreign body recognized is defined as on track
Equipment, extracts characteristics of image and is stored in wrong report feature database；
(6), under vehicle normal operation mode, in Dynamic Envelope, image block is taken from the near to the remote according to orbital path, will be away from car
Head the most nearby level altitude, a length of left rail to right gauge from orbital image block to be compared, with next-door neighbour orbital image to be compared
Block front level altitude, a length of left rail to right gauge from orbital image block carry out Characteristic Contrast, inconsistent preliminary of feature
Assert that orbital image block to be compared is foreign body intrusion outputting alarm；Again by its feature with wrong report feature database in characteristics of image by
One contrast, mates with any feature in wrong report feature database, then gets rid of the possibility of foreign body intrusion, do not mate, confirm as foreign body and invade
Limit.
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that described regard
Frequently acquisition module includes: be divided into the left and right sides, the first photographic head of synchronous acquisition video image in train head windshield
And second camera.
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that described step
Suddenly the preliminary treatment process of (2) is specific as follows:
The RGB image received is converted to gray level image by image pre-processing module；
Scharr operator is used to be filtered operation, filter coefficient
Image smoothing operates, and image carries out Gaussian convolution, and core size is 3*3, standard deviation sigma=(n/2-1) * 0.3+0.8,
The wherein corresponding horizontal core of n or vertical core size；
Smooth rear image is carried out secondary scaling: use Gaussian pyramid decomposition is to input picture down sample, the most right
Input picture Gaussian filter carries out convolution, then by refusing the row and column down sample image of even number；Use
Input picture to up-sampling, is first passed through and inserts 0 even number line and even column in the picture, then by Gaussian pyramid decomposition
The image Gaussian filter obtained is carried out Gaussian convolution, and its median filter is multiplied by 4 and does interpolation, and output image is input picture
4 times of sizes, Gaussian convolution core size is 5*5；
Eliminate picture noise, split independent pictorial element, use kernel structure be rectangle, size be the structural element of 3*3, sweep
Retouch each pixel of bianry image, do AND operation with the bianry image of structural element Yu covering, if being all 1, then structure chart
This pixel of picture is 1, is otherwise 0, makes bianry image reduce a circle, and this operation is in triplicate；Connect adjacent element, use kernel
Be shaped as rectangle, size is the structural element of 3*3, scans each pixel of bianry image, by the two-value of structural element Yu covering
Image does AND operation, if being all 0, then this pixel of structural images is 0, is otherwise 1, makes bianry image expand a circle, behaviour
Make three times；
Image is carried out binarization operation,Wherein thresh is 100.
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that described step
Suddenly the detailed process of (3) is: Hough transformation uses the built-in function HoughLinesP of opencv, and wherein rho is 1, and theta is pi/
180, threshold is 80；Left and right view uses SGBM algorithm calculate parallax and carries out three-dimensional reconstruction, then calculate two rectilineal intervals
From；Conversion formula between camera coordinate system and world coordinate systemWherein (Xc, Yc, Zc) represents P
Point position under camera coordinate system, (Xw, Yw, Zw) represents P point position under world coordinate system, and R is spin matrix, T
For translation matrix, the inside and outside parameter that R and T is demarcated by binocular camera obtains；Calculate two included angle of straight line, tan θ ＜ 0.1；Rim detection
Extracting profile, fit to curve, find the curve the highest with straight line degree of overlapping, smoothed curve obtains left and right orbital path.
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that step
Suddenly the characteristics of image in (5) is invariant moment features, comprises the combination after moment of the orign, central moment and normalization, two dimension
Image with f (x, y) represent, then its (p+q) rank moment of the orign is:Central moment is:Centre coordinate for image；Use zeroth order center
Remaining each rank central moment is normalized by square, obtains image normalization central momentWherein,Take p, q=0,1,2,3.
The method getting rid of wrong report on track the most according to claim 5 in foreign body intrusion identification, it is characterised in that described not
Bending moment uses seven kinds of not bending moment result of calculations, and the definition of seven kinds of not bending moments is:
m1=η20+η02
m3=(η30-3η12)2+(η03-3η21)2
m4=(η30+η12)2+(η03+η21)2
m5=(η30-3η12)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η03-3η21)(η03+η21)
m6=(η20-η02)[(η30+η12)2-(η21+η03)2]+4η11(η30+η12)(η03+η21)
m7=(3 η21-η03)(η30+η12)[(η30+η12)2-3(η21+η03)2]+(η30-3η12)(η21+η03)[(η21+η03)2-3(η30+
η12)2]。
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that step
(6) characteristic matching in is analog quantity tolerance,P is image to be identified, and q is in wrong report feature database
Image.
The method getting rid of wrong report on track the most according to claim 1 in foreign body intrusion identification, it is characterised in that wrong report spy
Levy stock to be stored in a hard disk, report feature database under vehicle normal operation mode by mistake and all read in internal memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610497533.4A CN106156780B (en) | 2016-06-29 | 2016-06-29 | The method of wrong report is excluded on track in foreign body intrusion identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610497533.4A CN106156780B (en) | 2016-06-29 | 2016-06-29 | The method of wrong report is excluded on track in foreign body intrusion identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106156780A true CN106156780A (en) | 2016-11-23 |
| CN106156780B CN106156780B (en) | 2019-08-13 |
Family
ID=57350344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610497533.4A Active CN106156780B (en) | 2016-06-29 | 2016-06-29 | The method of wrong report is excluded on track in foreign body intrusion identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106156780B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106585670A (en) * | 2016-12-09 | 2017-04-26 | 交控科技股份有限公司 | Video-based urban rail transit ahead train detection system and method |
| CN106991410A (en) * | 2017-04-17 | 2017-07-28 | 合肥酷庆信息科技有限公司 | A kind of railway foreign body intrusion distributed intelligent monitoring system based on machine vision |
| CN107144887A (en) * | 2017-03-14 | 2017-09-08 | 浙江大学 | A kind of track foreign body intrusion monitoring method based on machine vision |
| CN107909056A (en) * | 2017-11-30 | 2018-04-13 | 唐智科技湖南发展有限公司 | A kind of image processing method and its device of the identification of track foreign matter |
| CN108506170A (en) * | 2018-03-08 | 2018-09-07 | 上海扩博智能技术有限公司 | Fan blade detection method, system, equipment and storage medium |
| CN109117810A (en) * | 2018-08-24 | 2019-01-01 | 深圳市国脉畅行科技股份有限公司 | Fatigue driving behavioral value method, apparatus, computer equipment and storage medium |
| CN110217271A (en) * | 2019-05-30 | 2019-09-10 | 成都希格玛光电科技有限公司 | Fast railway based on image vision invades limit identification monitoring system and method |
| CN110619308A (en) * | 2019-09-18 | 2019-12-27 | 名创优品（横琴）企业管理有限公司 | Aisle sundry detection method, device, system and equipment |
| CN110654422A (en) * | 2019-11-12 | 2020-01-07 | 中科(徐州)人工智能研究院有限公司 | Rail train driving assistance method, device and system |
| CN111860385A (en) * | 2020-07-27 | 2020-10-30 | 珠海研果科技有限公司 | Target identification method and device |
| CN112092869A (en) * | 2020-09-23 | 2020-12-18 | 北京交大微联科技有限公司 | Crossing object identification method, crossing control method and protective equipment for rail transit |
| CN112542007A (en) * | 2020-11-30 | 2021-03-23 | 福州外语外贸学院 | Dangerous target detection method and system between financial withdrawals |
| CN113011252A (en) * | 2021-02-04 | 2021-06-22 | 成都希格玛光电科技有限公司 | Track foreign matter intrusion detection system and method |
| CN113390435A (en) * | 2021-05-13 | 2021-09-14 | 中铁二院工程集团有限责任公司 | High-speed railway multi-element auxiliary positioning system |
| CN113487518A (en) * | 2021-07-29 | 2021-10-08 | 中铁二十二局集团有限公司 | Method for detecting foreign matters in tunnel |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0376977A (en) * | 1989-08-17 | 1991-04-02 | Mitsubishi Electric Corp | Invation monitor system |
| CN102024145A (en) * | 2010-12-01 | 2011-04-20 | 五邑大学 | Layered recognition method and system for disguised face |
| CN103057569A (en) * | 2013-02-05 | 2013-04-24 | 中南大学 | Track traffic line foreign material invasion limit monitoring device and method |
| CN103996017A (en) * | 2014-02-24 | 2014-08-20 | 航天恒星科技有限公司 | Ship detection method based on Hu invariant moment and support vector machine |
-
2016
- 2016-06-29 CN CN201610497533.4A patent/CN106156780B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0376977A (en) * | 1989-08-17 | 1991-04-02 | Mitsubishi Electric Corp | Invation monitor system |
| CN102024145A (en) * | 2010-12-01 | 2011-04-20 | 五邑大学 | Layered recognition method and system for disguised face |
| CN103057569A (en) * | 2013-02-05 | 2013-04-24 | 中南大学 | Track traffic line foreign material invasion limit monitoring device and method |
| CN103996017A (en) * | 2014-02-24 | 2014-08-20 | 航天恒星科技有限公司 | Ship detection method based on Hu invariant moment and support vector machine |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106585670B (en) * | 2016-12-09 | 2018-04-17 | 交控科技股份有限公司 | To train detection systems and method before a kind of urban track traffic based on video |
| CN106585670A (en) * | 2016-12-09 | 2017-04-26 | 交控科技股份有限公司 | Video-based urban rail transit ahead train detection system and method |
| CN107144887A (en) * | 2017-03-14 | 2017-09-08 | 浙江大学 | A kind of track foreign body intrusion monitoring method based on machine vision |
| CN107144887B (en) * | 2017-03-14 | 2018-12-25 | 浙江大学 | A kind of track foreign body intrusion monitoring method based on machine vision |
| CN106991410A (en) * | 2017-04-17 | 2017-07-28 | 合肥酷庆信息科技有限公司 | A kind of railway foreign body intrusion distributed intelligent monitoring system based on machine vision |
| CN107909056B (en) * | 2017-11-30 | 2020-08-28 | 唐智科技湖南发展有限公司 | Image processing method and device for track foreign matter identification |
| CN107909056A (en) * | 2017-11-30 | 2018-04-13 | 唐智科技湖南发展有限公司 | A kind of image processing method and its device of the identification of track foreign matter |
| CN108506170A (en) * | 2018-03-08 | 2018-09-07 | 上海扩博智能技术有限公司 | Fan blade detection method, system, equipment and storage medium |
| CN109117810A (en) * | 2018-08-24 | 2019-01-01 | 深圳市国脉畅行科技股份有限公司 | Fatigue driving behavioral value method, apparatus, computer equipment and storage medium |
| CN110217271A (en) * | 2019-05-30 | 2019-09-10 | 成都希格玛光电科技有限公司 | Fast railway based on image vision invades limit identification monitoring system and method |
| CN110619308A (en) * | 2019-09-18 | 2019-12-27 | 名创优品（横琴）企业管理有限公司 | Aisle sundry detection method, device, system and equipment |
| CN110654422A (en) * | 2019-11-12 | 2020-01-07 | 中科(徐州)人工智能研究院有限公司 | Rail train driving assistance method, device and system |
| CN110654422B (en) * | 2019-11-12 | 2022-02-01 | 银河水滴科技（北京）有限公司 | Rail train driving assistance method, device and system |
| CN111860385A (en) * | 2020-07-27 | 2020-10-30 | 珠海研果科技有限公司 | Target identification method and device |
| CN112092869A (en) * | 2020-09-23 | 2020-12-18 | 北京交大微联科技有限公司 | Crossing object identification method, crossing control method and protective equipment for rail transit |
| CN112542007A (en) * | 2020-11-30 | 2021-03-23 | 福州外语外贸学院 | Dangerous target detection method and system between financial withdrawals |
| CN113011252A (en) * | 2021-02-04 | 2021-06-22 | 成都希格玛光电科技有限公司 | Track foreign matter intrusion detection system and method |
| CN113011252B (en) * | 2021-02-04 | 2023-12-05 | 成都希格玛光电科技有限公司 | Rail foreign matter intrusion detection system and method |
| CN113390435A (en) * | 2021-05-13 | 2021-09-14 | 中铁二院工程集团有限责任公司 | High-speed railway multi-element auxiliary positioning system |
| CN113390435B (en) * | 2021-05-13 | 2022-08-26 | 中铁二院工程集团有限责任公司 | High-speed railway multi-element auxiliary positioning system |
| CN113487518A (en) * | 2021-07-29 | 2021-10-08 | 中铁二十二局集团有限公司 | Method for detecting foreign matters in tunnel |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106156780B (en) | 2019-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106156780A (en) | The method getting rid of wrong report on track in foreign body intrusion identification | |
| CN104573646A (en) | Detection method and system, based on laser radar and binocular camera, for pedestrian in front of vehicle | |
| US9269001B2 (en) | Illumination invariant and robust apparatus and method for detecting and recognizing various traffic signs | |
| CN104029680B (en) | Lane Departure Warning System based on monocular cam and method | |
| Zhou et al. | LIDAR and vision-based real-time traffic sign detection and recognition algorithm for intelligent vehicle | |
| CN105260712B (en) | A kind of vehicle front pedestrian detection method and system | |
| CN103971097B (en) | Vehicle license plate recognition method and system based on multiscale stroke models | |
| US20150339533A1 (en) | Method and apparatus for detecting target object in blind area of vehicle | |
| CN106156725A (en) | A kind of method of work of the identification early warning system of pedestrian based on vehicle front and cyclist | |
| Li et al. | Nighttime lane markings recognition based on Canny detection and Hough transform | |
| CN103077526B (en) | There is train method for detecting abnormality and the system of depth detection function | |
| CN103971087B (en) | Method and device for searching and recognizing traffic signs in real time | |
| CN103824452A (en) | Lightweight peccancy parking detection device based on full view vision | |
| CN109506664A (en) | Device and method is provided using the guidance information of crossing recognition result | |
| CN105933611A (en) | Dynamic envelope line drawing method for preventing camera jitter in rail foreign matter invasion recognition | |
| CN103679205B (en) | Assume based on shade and the Foregut fermenters method of layering HOG symmetrical feature checking | |
| Xu et al. | Real-time obstacle detection over rails using deep convolutional neural network | |
| CN106529530A (en) | Monocular vision-based ahead vehicle detection method | |
| US20140152821A1 (en) | Pedestrian detection systems and methods | |
| CN106250816A (en) | A kind of Lane detection method and system based on dual camera | |
| CN104881661B (en) | Vehicle checking method based on structural similarity | |
| JP5077088B2 (en) | Image processing apparatus and image processing method | |
| JP4826306B2 (en) | Facial part tracking method and apparatus | |
| CN104574335A (en) | Infrared and visible image fusion method based on saliency map and interest point convex hulls | |
| CN103824060A (en) | Method for extracting fingerprint detail points |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |