CN106023076B - The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch - Google Patents
The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch Download PDFInfo
- Publication number
- CN106023076B CN106023076B CN201610312137.XA CN201610312137A CN106023076B CN 106023076 B CN106023076 B CN 106023076B CN 201610312137 A CN201610312137 A CN 201610312137A CN 106023076 B CN106023076 B CN 106023076B
- Authority
- CN
- China
- Prior art keywords
- protective fence
- panorama sketch
- guardrail
- video
- pixel
- 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
- 230000001681 protective effect Effects 0.000 title claims abstract description 103
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000001514 detection method Methods 0.000 title claims abstract description 53
- 230000006378 damage Effects 0.000 title claims abstract description 15
- 238000005304 joining Methods 0.000 title description 4
- 230000007547 defect Effects 0.000 claims abstract description 21
- 238000013442 quality metrics Methods 0.000 claims abstract description 4
- 238000005070 sampling Methods 0.000 claims description 34
- 230000011218 segmentation Effects 0.000 claims description 26
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 9
- 229910052742 iron Inorganic materials 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 description 15
- 230000033001 locomotion Effects 0.000 description 9
- 238000000605 extraction Methods 0.000 description 8
- 230000003287 optical effect Effects 0.000 description 7
- 238000012360 testing method Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 238000007689 inspection Methods 0.000 description 5
- 230000003068 static effect Effects 0.000 description 5
- 230000004888 barrier function Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 239000004568 cement Substances 0.000 description 3
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000000926 separation method Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 238000011410 subtraction method Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000009194 climbing Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000001808 coupling effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30221—Sports video; Sports image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention provides a kind of methods of Panoramagram montage method based on virtual sampling channel model and the damage condition of the protective fence of the detection high-speed railway of Panoramagram montage method.The detection method, comprising: obtain the propulsion video of record high-speed railway ambient condition；It is the left and right panorama sketch of railway operation environment by the propulsion Video Quality Metric, as protective fence panorama sketch based on the Panoramagram montage method of virtual sampling channel model；Divide principle according to entropy is maximized, the position of guardrail and background in the protective fence panorama sketch is positioned, obtains the two-value code of 0 and 1 composition；Run-length encoding is carried out to the two-value code；According to the run-length encoding, the pixel wide D of the background area between adjacent two guardrail is calculatedcur；According to the pixel wide D of the background area between adjacent two guardrailcurWith the normal pixel spacing d between adjacent two guardrail, judge adjacent two guardrail with the presence or absence of defect.
Description
Technical field
The present invention relates to computer application technology more particularly to a kind of joining methods of panorama sketch and detection high-speed iron
The method of the damage condition of the protective fence on road.
Background technique
Safety is basis and the premise of Development of High Speed Railway.There are many factor for influencing high-speed railway safety, are related to people, column
Vehicle, track and environment and the coupling effect between them.In order to guarantee high-speed railway traffic safety, it is desirable that high-speed railway
Running environment must realize whole closing.It is a kind of means for realizing enclosed environment that line of high-speed railway, which installs protective fence additional,.
It is invaded within protective fence to monitor artificial destruction or climbing, and signalling arrangement and power supply in monitoring enclosed environment
The fine status of the infrastructure such as system, currently, being mainly monitored using fixed video equipment.For example, in Beijing-Shanghai high speed
On railway, more than 400 video monitoring equipments are mounted on Beijing-Shanghai High-Speed Railway roadbed, crossing, bridge, public affairs across iron, bottle-neck section etc.
Position guarantees vehicle safety operation.
In common video monitoring system, picture pick-up device is in fixed static state, and monitored object is likely to be at fortune
Dynamic state.The situation that target is detected using the relative motion of target and background in video sequence is a kind of common moving target
Detection method, wherein mainly having: the detection methods such as background subtraction method, time differencing method and optical flow method (Optical flow).It should
The advantages of class method, without training, can be detected online；And the disadvantage is that respective objects can only be searched out, and can not sentence
Bright detection zone is which kind of specific target, and the later period is needed further to judge.
Background subtraction method depends primarily on the accuracy to background image modeling, but in the actual environment, due to environment
The factors such as variation influence, and exacerbate the extraction of background and the difficulty updated, increase difficulty for accurate target of extracting.Therefore, mesh
Such preceding method mostly uses greatly statistical learning method to analyze successive video frames, background modeling, and over time
Then online updating background model detects moving region by the difference between present frame and background model again.
Time difference method is a kind of relatively easy, lesser method of operand, by the front and back two for calculating video sequence
The pixel difference of frame corresponding position, if more than the threshold value of setting, then it is assumed that be object pixel, be otherwise background pixel.Optical flow method is adopted
With the vector of moving target moving target can be detected with the characteristic of time change.The advantages of this method is even if in video camera
Also it can detecte out moving target in the case where movement.Its major defect: it is more sensitive to noise, it is computationally intensive, be not suitable for
The higher occasion of requirement of real-time.
But fixed point monitor mode is limited by the acquisition visual field, cannot control route and the along the line whole circumstances, because
This monitors front side line environment and along both side using the vehicle end surroundings monitoring apparatus for being mounted on high speed comprehensive detection train
Protective fence state is a kind of effective method.
The Detection task of influence train operating safety is completed using dedicated high speed comprehensive detection train in the world.Detection
Content generally comprises: contact net geometry, contact line abrasion, bow net effect, electric parameter, gauge, track geometry, rail profile and
Undulatory wear, car body and axle box acceleration, wheel-rail force, track and bogie, communication check and positioning etc..In addition to this,
High speed comprehensive detection train both domestic and external is equipped with the video equipment as automobile data recorder in its front and rear, is transported with forward direction
The mode of dynamic video is used to obtain circumstance state information on the way, whether there is abnormal shape for later period artificial detection Along Railway environment
State provides foundation.How to influence to set in environment closure and foreign matter or route from quick obtaining in the multitude of video image of acquisition
Standby exception invades the information of limit, and carries out correct early warning, is high-speed railway urgent problem to be solved.
Summary of the invention
The embodiment provides the defects of a kind of joining method of panorama sketch and the protective fence of detection high-speed railway
The video data of magnanimity is carried out lossless information extraction with the panorama bitmap-format of lightweight, reduces video by the method for state
Data mode stores and accesses expense.
To achieve the goals above, this invention takes following technical solutions:
A kind of Panoramagram montage method based on virtual sampling channel model, comprising:
Obtain the propulsion video of record high-speed railway ambient condition；Frame number is extracted from the propulsion video is
The sequence of video images of N；
According to the railway scene structure determined via end point, external sampling rectangle OR is set in every frame imagem；
According to the speed of train, internal sample rectangle IR is set in every frame imagem；
It will be by the external sampling rectangle OR of every frame imagemWith the internal sample rectangle IRmThe rectangular ring of composition is adopted
Sample annulus is divided into four strips mosaic region St,Sb,Sl,Sr；
By four strips mosaic regions S of every frame imaget,Sb,Sl,SrBy image volume around, be corrected to rule square
Shape strips St',Sb',Sl',Sr'；
By 4 × the N number of corrected shape of rectangular ribbon St',Sb',Sl',Sr', it is carried out respectively according to respective sample plane
Split generates the panorama sketch of 4 planes of railway scene.
A method of the damage condition of the protective fence of the detection high-speed railway based on the Panoramagram montage method, packet
It includes:
Step 1 obtains the propulsion video of record high-speed railway ambient condition；
The propulsion Video Quality Metric is by step 2 based on the Panoramagram montage method of virtual sampling channel model
The left and right panorama sketch of railway operation environment, as protective fence panorama sketch；
Step 3 divides principle according to entropy is maximized, to the position of guardrail and background in the protective fence panorama sketch into
Row positions, and indicates the position of protective fence, the position of background to be indicated at F (j)=0 at F (j)=1, obtains the two-value of 0 and 1 composition
Code；J is the column serial number of protective fence panorama sketch；
Step 4 carries out run-length encoding to the two-value code；
Step 5 calculates the pixel wide D of the background area between adjacent two guardrail according to the run-length encodingcur；
Step 6, according to the pixel wide D of the background area between adjacent two guardrailcurWith adjacent two guardrail
Between normal pixel spacing d, judge that adjacent two guardrail with the presence or absence of defect, generates judging result；
Step 7 exports the judging result.
As can be seen from the technical scheme provided by the above-mentioned embodiment of the present invention, in the embodiment of the present invention, by the view of magnanimity
The panorama bitmap-format of frequency lightweight accordingly carries out lossless information extraction, reduces storing and accessing out for video data form
Pin.
The additional aspect of the present invention and advantage will be set forth in part in the description, these will become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill of field, without any creative labor, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 is the flow diagram of the Panoramagram montage method of the present invention based on virtual sampling channel model；
Fig. 2 is the damage condition of the protective fence of the detection high-speed railway of the present invention based on Panoramagram montage method
Method；
Fig. 3 is high speed railway scene sampling ring geometry of the present invention；
Fig. 4 is the image registration in the present invention between adjacent two frame；
The histogram distribution that Fig. 5 is the three-dimensional feature F (j) of pixel column in panorama sketch in the present invention；
Fig. 6 is in the present invention based on Mean-Variance-gradient Threshold segmentation；
Fig. 7 is that the run-length encoding of guardrail panorama sketch in the present invention indicates；
Fig. 8 is the panorama sketch generated from closing Railway Environment in the present invention: (a) left side panorama sketch；(b) right side panorama
Figure；(c) bottom panorama sketch；(d) top panorama sketch
Fig. 9 is the video image of comprehensive detection train acquisition in the present invention: (a) guardrail N/D；(b) guardrail has defect
Figure 10 is part railway panorama sketch in the present invention: (a) panorama sketch on the left of railway；(b) protection intercepted from (a)
Column panorama sketch
Figure 11 is the binarization segmentation result of guardrail panorama sketch in the present invention: (a) gray scale one dimensional histograms maximum entropy threshold
Segmentation；(b) Gray Level-Gradient (GLGM) two-dimensional histogram maximum entropy threshold is divided；(c) combination gray average-proposed by the present invention
The segmentation of variance-gradient stereogram maximum entropy threshold
Figure 12 is the binarization segmentation result of high-speed rail guardrail panorama sketch in the present invention: (a) the guardrail panorama sketch of high-speed railway；
(b) gray scale one dimensional histograms maximum entropy threshold is divided；(c) Gray Level-Gradient (GLGM) two-dimensional histogram maximum entropy threshold is divided；
(d) combination gray average proposed by the present invention-variance-gradient stereogram maximum entropy threshold segmentation
Figure 13 is high speed railway protective of the present invention column defect rapid detection method flow chart of steps.
Specific embodiment
Embodiments of the present invention are described below in detail, the example of the embodiment is shown in the accompanying drawings, wherein from beginning
Same or similar element or element with the same or similar functions are indicated to same or similar label eventually.Below by ginseng
The embodiment for examining attached drawing description is exemplary, and for explaining only the invention, and is not construed as limiting the claims.
It as described in Figure 1, is a kind of Panoramagram montage method based on virtual sampling channel model of the present invention, packet
It includes:
Step 11, the propulsion video of record high-speed railway ambient condition is obtained；It is mentioned from the propulsion video
Taking frame number is the sequence of video images of N；
Step 12, according to the railway scene structure determined via end point, external sampling rectangle is set in every frame image
ORm；
Step 13, according to the speed of train, internal sample rectangle IR is set in every frame imagem；
It step 14, will be by the external sampling rectangle OR of every frame imagemWith the internal sample rectangle IRmThe square of composition
Shape ring-type sampling ring band is divided into four strips mosaic region St,Sb,Sl,Sr；
Step 15, by four strips mosaic regions S of every frame imaget,Sb,Sl,SrBy image volume around being corrected to
The shape of rectangular ribbon S of rulet',Sb',Sl',Sr'；
Step 16, by 4 × the N number of corrected shape of rectangular ribbon St',Sb',Sl',Sr', according to respective sample plane
Split is carried out respectively, generates the panorama sketch of 4 planes of railway scene.
The step of determining the end point specifically:
Coordinate of the end point in image coordinate system is (x0,y0)T；The analytic expression of i-th line section is x+kiy+bi=0, ki
For the slope of i-th line section, biFor intercept；Weight wiIt is the length of i-th line section；N is total column of pixel in protective fence panorama sketch
Number；
Wherein,
In the embodiment of the present invention, the video data of magnanimity is subjected to lossless information with the panorama bitmap-format of lightweight and is taken out
It takes, reduce video data form stores and accesses expense, can be used for subsequent processing.
As described in Figure 2, the protective fence of the detection high-speed railway to be of the present invention based on Panoramagram montage method is scarce
The method of damage state, comprising:
Step 21, the propulsion video of record high-speed railway ambient condition is obtained；
Step 22, the propulsion Video Quality Metric is by the Panoramagram montage method based on virtual sampling channel model
The left and right panorama sketch of railway operation environment, as protective fence panorama sketch；
Step 23, divide principle according to maximizing entropy, to the position of guardrail and background in the protective fence panorama sketch into
Row positions, and indicates the position of protective fence, the position of background to be indicated at F (j)=0 at F (j)=1, obtains the two-value of 0 and 1 composition
Code；J is the column serial number of protective fence panorama sketch；
Step 24, run-length encoding is carried out to the two-value code；
Step 25, according to the run-length encoding, the pixel wide D of the background area between adjacent two guardrail is calculatedcur；
Step 26, according to the pixel wide D of the background area between adjacent two guardrailcurWith adjacent two guardrail
Between normal pixel spacing d, judge that adjacent two guardrail with the presence or absence of defect, generates judging result；Wherein, the step
26 include: to judge DcurWhether kd is greater than, wherein k is experience adjustment parameter.
Step 27, the judging result is exported.
In the embodiment of the present invention, the video data of magnanimity is subjected to lossless information with the panorama bitmap-format of lightweight and is taken out
Take, reduce video data form stores and accesses expense, moreover, by video be converted into it is a kind of more suitable for manually inspect with
And the form of computer analysis processing, it solves and is quickly converted by the undistorted of propulsion video to panorama sketch.
The step 22 includes:
The sequence of video images that frame number is N is extracted from the propulsion video；
According to the railway scene structure determined via end point, external sampling rectangle OR is set in every frame imagem；
According to the speed of train, internal sample rectangle IR is set in every frame imagem；
By every frame image by the external sampling rectangle ORmWith the internal sample rectangle IRmThe rectangular ring of composition is adopted
Sample annulus is divided into four strips mosaic region St,Sb,Sl,Sr；Wherein, St,Sb,Sl,SrRespectively sky, rail, left side shield
Column and right side guardrail region；
By four strips mosaic regions S of every frame imaget,Sb,Sl,SrBy image volume around, be corrected to rule square
Shape strips St',Sb',Sl',Sr'；
By 2 × the N number of corrected shape of rectangular ribbon Sl',Sr', split is carried out respectively according to respective sample plane, it is raw
At the left and right panorama sketch of railway scene.
The step of determining the end point specifically:
Coordinate of the end point in image coordinate system is (x0,y0)T；The analytic expression of i-th line section is x+kiy+bi=0, ki
For the slope of i-th line section, biFor intercept；Weight wiIt is the length of i-th line section；N is total column of pixel in protective fence panorama sketch
Number；
Wherein,
The step 23 includes:
For each column j of protective fence panorama sketch, gray average M (1, j), standard deviation V (1, j) and gradient mean value G are calculated
(1,j)；J=1 ..., N；N is total columns of pixel in protective fence panorama sketch；
Using F (j) maximum entropy, segmentation threshold (ε is calculated*,ξ*,η*)；F (j) is perpendicular for what is extracted in the protective fence panorama sketch
The feature distribution of the upward each column pixel of histogram；
According to the segmentation threshold (ε*,ξ*,η*), binarization segmentation is carried out to F (j), protective fence region is set to 1, background
Region is set to 0；
Wherein,；V (1, j) is the gray average of each column pixel in protective fence panorama sketch；
M (1, j) is the gray variance of each column pixel in protective fence panorama sketch；G (1, j) is each column picture in protective fence panorama sketch
The gradient mean value of element.
Wherein, p (i, j) is the gray value of pixel (i, j) in protective fence panorama sketch, th,bhThe respectively top of protective fence
With coordinate position of the bottom in protective fence panorama sketch.
Another embodiment of the present invention is described below.
The present invention provides a kind of utilization propulsion view using the closure for detecting high-speed railway closed operation environment as object
The Panoramagram montage of frequency, and detected by panorama sketch feature, realize the dynamic detection of high-speed railway protective fence defect.
The technical scheme is that: the propulsion video for recording high-speed railway ambient condition is sampled by virtual channel
Splicing is converted to Railway Environment panorama sketch；Feature extraction is carried out to panorama sketch；According to the feature row obtained after feature extraction
Journey coding provides the judgement of railway protective column damage condition and assessment.That is, video image is transformed to a kind of static panorama
Bitmap-format generates propulsion video panorama under camera optical axis and direction of motion uniform condition；It is several based on detection zone
What virtual sampling channel model of structural generation propulsion video；Mean-Variance-gradient (MVG) based on protective fence panorama sketch
Stereogram maximum entropy threshold divides and the guardrail detection based on run-length encoding.
Comprising the following steps:
Step 1, propulsion video is positioned based on rail track priori, the step of virtual rectangle sampling channel；
Step 2, segmentation propulsion video frame be day, left and right four parts of images the step of；
Step 3, the step of Panoramagram montage based on propulsion video frame calculates；
Step 4, to protective fence panorama sketch feature calculation the step of；
Step 5, the guardrail detecting step based on run-length encoding.
Wherein, step 1 specifically includes:
One group of annulus sequence { S is extracted from video sequence1,S2,S3..., SN, and meet adjacent overall view ring belt to object
It manages and is neither overlapped again without interval, i.e., " fully sampled " between the sampling in space.Splicing exists after annulus sequence is carried out geometric corrections
Panoramic picture is generated together.
Based on rail track priori, propulsion video virtual channel is positioned, is constructed with the object of depth layer of interest
The virtual rectangle sampling channel of " fully sampled ".The building of virtual rectangle sampling channel is based on following objective approximation:
(1) rail facility scene includes: the protective fence or sound barrier of two sides, the rail on ground and is suspended on connecing for top
Touch net etc., similar facility, which is all considered as, is located approximately at same plane, and known to the distance of plane separation camera；
(2) each plane of rail facility scene is the depth layer nearest apart from camera, in the panorama sketch of generation there is no because
Rail facility loss of learning caused by lack sampling.
Step 2 specifically includes:
(1) end point is determined
The video image of video sequence is examined using LSD (line segment detector) Line Segment Detection Algorithm
It looks into, all line segments detected is divided into two groups, first group is the line segment with trunnion axis angle in [- 60 °, 60 °] range, is used
To use Least Square Method to be parallel to the end point that the line segment of rail is formed；Second group for vertical axes angle [- 5 °,
5 °] line segment in range, for estimating the position of line bar, to divide railway scene.
If coordinate of the end point in image coordinate system is (x0,y0)T, it and i-th line section x+kiy+bi=0 (kiIt is oblique
Rate, biFor intercept) distance be
WhereinAccording to least square method, point (x is minimized0,y0)TWith all line segments spacing
From quadratic sum, i.e.,
Wherein, weight wiIt is the length of i-th line section.
X is asked respectively to formula (2)0And y0Derivative, and enabling it is zero, obtains following system of linear equations,
Here
N is total columns of pixel in guardrail panorama sketch；
The solution of formula (3) is,
(x0,y0) be exactly end point optimal estimation.In order to improve the accuracy of estimating vanishing point, using cross-iteration
Method rejects the line segment of " mistake ".
(2) divide day, left and right four part.
The virtual rectangle sampling channel rectangular parallelepiped structure being made of rail, contact net and protective fence or sound barrier has solid
Fixed geometry.The shape of virtual rectangle sampling channel according to scene determines that (Fig. 3 is high-speed iron to selected sampling ring position
The scene sample ring on road).
It is separately connected end point Q to rectangular four vertex of sampling ring, just by four parts of this four wire segmentings
Be day, left and right four parts.
Step 3. is calculated based on the Panoramagram montage of propulsion video frame.
As shown in figure 4, as follows based on vanishing Point Detection Method and camera motion information image registration process:
(1) by the railway scene of concern, i.e., the position where guardrail, contact net, sound barrier, rail constitutes a virtual square
Shape sampling channel, and " fully sampled " is carried out in virtual rectangle sampling channel.
(2) sequence of video images { I that one section of frame number is N is given1,I2..., IN-1, IN, splicing is chosen in every frame image
Region determines outer rectangular ORm:AmBmCmDmAnd inner rectangular IRm:A′mB′mC′mD′m。
(3) scene structure determined according to end point preassigns outer rectangular ORm:AmBmCmDmPosition, and in each frame
It is the location of middle all identical.Four are drawn respectively by the top and bottom of electric pole at left and right sides of image by end point Q
Ray, this four rays by scene cut be sky, rail, four parts of left side guardrail and right side guardrail (day i.e. above,
Ground, left and right four parts).Four vertex A of outer rectangularm、Bm、Cm、DmIt should respectively fall on four rays, it thus can be true
The panorama sketch content for protecting four scene of upper and lower, left and right generated does not overlap, and avoids wrong sampling, while also simplifying virtual field
The process that scape is drawn.Furthermore since the direction of motion of video camera is parallel with rail, in image in the expansion of pixel light stream
The heart (Focus of Expansion, FOE) is overlapped with the end point Q that two rail are formed at a distance in image.Square external in this way
The moving direction on four vertex of shape, i.e. outer rectangular ORmAs the zoom direction of time frame is just uniquely determined through end point Q,
It is denoted as:
(4) train speed is set as V, and the frame speed of video camera is R, then the distance of spatial sampling is V/R between adjacent two frame.If
For the inner parameter of camera it is known that then being calculated by corresponding geometry, the movement speed i.e. image that can acquire pixel in image is fast
Spend v.
(5) outer rectangular OR in m frame is obtainedm:AmBmCmDmZoom directionAnd image
After speed v, corresponding coordinate position can be by m-1 frame for the outer rectangularDirectly acquire.
Outer rectangular OR in (6) m-1 framesm-1With inner rectangular IRm-1Constitute the splicing regions S of rectangular ringm, will
SmAlong the zoom direction on outer rectangular vertexIt is divided into 4 pieces of trapezoidal strips Sst、Sb、Sl、Sr, respectively day
Sky, rail, left side guardrail and right side guardrail region.
(7) using image volume around (Image Wrapping), to the strips S of constructiont、Sb、Sl、SrGeometric transformation is carried out, from
And irregular trapezoid-shaped strips are mapped as to the shape of rectangular ribbon S of rulet'、Sb'、Sl'、Sr'.The above mistake is repeated to every frame image
Journey obtains four groups of band sequences, these sequences is successively stitched together respectively and have just obtained panorama sketch.The present invention relates to left and right
Shape of rectangular ribbon S is only used in the detection of protective fence thusl'、Sr' splice the obtained left and right panorama sketch of railway operation environment, that is, it protects
Column panorama sketch.
Step 4. protective fence panorama sketch feature calculation.
Enabling p (i, j) is the gray value of pixel (i, j) in protective fence panorama sketch, th,bhRespectively the top of protective fence and
Coordinate position of the bottom in panorama sketch, then the gray average, gray variance of each column pixel and gradient mean value are denoted as respectively
After carrying out gray scale and Gradient Features extraction along protective fence panorama sketch vertical direction, obtain
{ F (j) }={ (x, y) }={ (M (1, j), V (1, j), G (1, j)) } (9)
Here 1≤j≤N, N are total columns of pixel in protective fence panorama sketch, and F (j) indicates the three-dimensional feature of each column pixel
Distribution.
Gray-scale statistical characteristics at protective fence meet: M (1, j) < ε, V (1, j) < ξ, G (1, j) < η.Here ε, ξ and η are to use
Foreground area in divided shielding column is the segmentation threshold for positioning protective fence position.
It is as shown in Figure 5 from the feature distribution F (j) of each column pixel on the vertical direction extracted in protective fence panorama sketch.
It is to be split to the feature distribution F (j) of guardrail protective fence, then three-dimensional if giving any one threshold value (ε, ξ, η)
The spatial distribution of feature F (j) is divided into 8 characteristic sub-areas (as shown in Figure 6) by (ε, ξ, η), is denoted as:
Each protective fence has approximate brightness value and uniform Luminance Distribution in the vertical direction, and background area is then
The scene of some brightness irregular distributions.Which results in the features of protective fence position to be mostly focused on R1Region, and the spy of background
Sign is mostly focused on R8Region.Although other 6 regions also contain the guardrail and background information of part, calculated to simplify
Process, ignore these secondary feature regions the experimental results showed that having no big influence to positioning result.
Any one feature vector is located at guardrail region R in F (j)1Or background area R8Probability be respectively PF(ε,ξ,
η) and PB(ε,ξ,η)；
Here, pxyzIt is located at the probability in prospect or background for the pixel in image.
And meet
PF(ε,ξ,η)+PB(ε,ξ,η)≈1 (13)
According to the definition of entropy, the three-dimensional entropy of guardrail region and background area is respectively
The general three entropy of F (j) is
H (ε, ξ, η)=HF(ε,ξ,η)+HB(ε,ξ,η) (16)
Principle is divided by maximum entropy, above formula H (ε, ξ, η) can be made to obtain the triple (ε of maximum value*,ξ*,η*) it is exactly institute
The optimal segmenting threshold asked, i.e.,
According to threshold value (ε*,ξ*,η*) binarization segmentation is carried out to F (j), protective fence region is set to 1, background area is set to
0, i.e.,
Step 5. is detected based on the guardrail of run-length encoding
Divide principle according to entropy is maximized, can by railway scene guardrail and background positions position, indicate at F (j)=1
The position of protective fence indicates the position of background at F (j)=0.As shown in figure 5, white area (value is 1) indicates protective fence, and it is black
Color region (value is 0) indicates background.Stroke is carried out to the two-value code of a succession of 0 and 1 composition obtained according to threshold segmentation method
Coding.
If guardrail is there are defect, the pixel wide D of the background area between adjacent two protective fencecurIt is naturally larger than protection
Normal pixel spacing d between column, if therefore Dcur> kd (k is experience adjustment parameter), then can be determined that the region exists and lack
Damage.
Another embodiment of the present invention is described below.
A kind of embodiment 1: new Panoramagram montage method based on propulsion video virtual rectangle sampling channel model.
The acquisition of railway panorama sketch includes acquisition, construction splicing regions, band spelling the following three steps: forward video
It connects.
It is generated shown in the following algorithm 1 of algorithm of railway scene panorama sketch according to virtual rectangle sampling channel model.
As shown in fig. 6, the panorama sketch knot generated for the low-quality video (720 × 576) that (150km/h) under high-speed condition is acquired
Fruit.Satisfactory panoramic picture can be generated in method proposed by the present invention.Protective fence and electric pole such as nearby does not lose letter
It ceases and distortion is smaller.The electric pole of distant place is apparent since " over-sampling " has occurred to stretch distortion, but this is not in actually detected
The part being concerned about.
Embodiment 2: the guardrail detection based on Panoramagram montage.
As shown in figure 9, for the propulsion video image of comprehensive detection train acquisition.Train keeps relatively at the uniform velocity camera shooting
The acquisition frame rate of machine is 25 frames/second.Fig. 9 (a) is the railway scene of protective fence N/D, and Fig. 9 (b) is that there are the rail yards of defect
Scape.
For the part railway panorama sketch (left side) of generation such as shown in Figure 10 (a), Figure 10 (b) is the portion extracted from Figure 10 (a)
Divide protective fence panorama sketch (i ∈ [405,490], j ∈ [2500,3000], wherein i is the row coordinate of pixel, and j is column coordinate).
According to maximizing, the protective fence protective fence location algorithm that entropy segmentation principle is realized is as follows:
Figure 11 be stereogram proposed by the present invention maximize entropy split plot design and one-dimensional grey level histogram split plot design and
Two dimensional gray-histogram of gradients split plot design comparing result.Figure 11 (a) is according only to gray scale one dimensional histograms maximum entropy threshold point
It is cutting as a result, it can be observed that many background areas all by erroneous segmentation be protective fence, this will lead to detection and fails to report.Figure 11 (b)
For according to Gray Level-Gradient (GLGM) two-dimensional histogram maximum entropy threshold divide as a result, it can be observed that 2770 position of column coordinate
The protective fence at place is background by erroneous segmentation, this will lead to detection wrong report.Figure 11 (c) is based on three-dimensional histogram proposed by the present invention
Figure segmentation result, by with Figure 10 (b) shown in compared with guardrail physical location, it can be seen that the knot of stereogram segmentation
Fruit is the most accurate.
After extracting the position of guardrail, run-length encoding is carried out to two-value code, to reach the lightweight storage of guardrail information
With access purpose.Quickly identify the damage location of protective fence from guardrail panorama sketch run-length encoding using algorithm 3 later.Such as Figure 11
(c) shown in, there are the region of apparent guardrail defect at two in the pixel coverage of [2500,3500], it is located at section
[3000,3138] and at [3290,3462].
Experiment video is avi format, and Shi Changyue 1 hour, data volume 1.35GB, sum of all pixels was (1024 × 768)/frame
× 85500 frames.The image for the jpg format that data volume is 84MB is generated using panoramic mosaic, sum of all pixels is 267500 × 600
(sum of all pixels × vertical direction sum of all pixels of horizontal direction).List is only used into the foreground and background region of guardrail in panorama sketch
The number of byte indicates that finally obtain 8917 position digital codings (37 3 41 3 33 2 42 3 ...), data volume is only
17.4KB.Therefore, --- MB grades --- KB grades from GB grades of data volume of the multi-stage compression that the present invention realizes railway guardrail is extracted,
Overcome because the video data volume is big, informationm storage and retrieval it is difficult and caused by detection algorithm apply bottleneck.
The present invention gives the algorithms of the defect location based on panorama sketch run-length encoding detection protective fence protective fence.
Since the traveling stroke of train is very long, the illumination of entire protective fence panorama sketch is because by weather, environment, the factors such as direction
Influence will appear variation, therefore, it is necessary to which protective fence panorama sketch is divided into K subgraph, and calculate one properly for every width subgraph
Local threshold (εi,ξi,ηi), i=1,2,3 ..., K here.
Detection is repeatedly manually visualized by live, defect at the guardrail physical presence 38 in this section of route is determined, is denoted as GT
=38.Testing result comparison based on different threshold segmentation methods is as shown in table 1, by comparing with the real conditions of guardrail defect
It was found that method proposed by the present invention has correctly detected defect at 36 (TP=36)；There is at 3 normal protective fence position by mistake
Inspection be defect (FP=3)；There is the breakage missed at two not to be detected (FN=2).Used here as Detection accuracy and recall
Rate verifies the validity of detection method, is achieved based on threshold segmentation method proposed by the present invention in testing result comparison
92.3% accuracy rate and 94.7% recall rate, better than the testing result of other partitioning algorithms.
The testing result comparison of the different threshold segmentation methods of table 1
In order to further verify the applicability of the maximum entropy threshold dividing method in conjunction with multiple features MVG stereogram, I
Again to high-speed railway another kind of protective fence --- white cement protective fence is similarly tested.The cement protection of white
Column is different from the irony guardrail of the dimness of color shown in Figure 10 (b), therefore under normal circumstances, the brightness in protective fence region is instead
It is higher than background area.As shown in Figure 12 (d), be high-speed railway protective fence panorama sketch segmentation result, by with figure
Protective fence panorama sketch in 12 (a) compares, it can be found that partitioning algorithm proposed by the present invention is either directed to metal protection column
Or cement protective fence, all achieves satisfactory segmentation result.
Figure 13 is the step flow chart of high-speed railway protective fence defect rapid detection method.
Wherein, a kind of high-speed railway protective fence detection method based on propulsion video panorama, comprising the following steps:
Propulsion is regarded after positioning propulsion video virtual rectangle sampling channel position according to rail track priori
Frequency frame be divided into day, left and right four part, and respectively to four parts carry out Panoramagram montage.That is, with panorama map grid
Video data is carried out lossless information extraction by formula, has obtained lightweight detection image information, is not only dropped relative to video data
It is low to store and access expense, and convert video to a kind of more suitable for artificial inspection or the shape of computer analysis processing
Formula.
Panoramagram montage based on propulsion video frame calculates, by the geometry priori of scene and the movement of camera
Velocity information quickly obtains the position of sampling ring, realizes the quick alignment between stitching image, generates left and right protective fence panorama
Figure, whole display effect are substantially better than the L-K optical flow method of characteristic matching.That is, virtually being adopted using propulsion video
Sample channel pattern overcomes the static fuzzy of propulsion video panorama narrowband splicing, and utilizes the elder generation of region geometry structure
It tests, constructs the panoramic mosaic algorithm of quick propulsion video.
Left and right protective fence panorama sketch is split using the maximum entropy threshold dividing method of MVG stereogram, is divided
Out in panorama sketch guardrail foreground and background region, realize protective fence position automatic positioning.To extracting protective fence position two
It is worth code and carries out run-length encoding, the lightweight storage and access of guardrail information is realized, according to algorithm 3 from the stroke of guardrail panorama sketch
The position on damage protection column is quickly identified in coding.
That is, the automated detection method of the guardrail defect based on guardrail panorama sketch (Fence panorama), benefit
The position that vertical protective fence in guardrail is automatically extracted with the maximum entropy threshold dividing method based on MVG stereogram, make its with
Background image separation；And compress the two-value code after the separation of guardrail panorama sketch with run-length encoding, the coded format of compression contains
Whole location informations of guardrail, corresponding decoding algorithm restore the position of guardrail from coding, and realize the missing inspection of guardrail
It surveys.
The invention has the following advantages:
(1) the new Panoramagram montage method of one kind proposed by the present invention, this method are a kind of empty based on propulsion video
The Panoramagram montage method of quasi- sampling channel model, the propulsion filled up under camera optical axis and direction of motion uniform condition regard
The blank of frequency panorama picture formation；Can be widely applied to the automatic Pilots of the delivery vehicles such as rail traffic, highway, safety inspection,
The detection of driving recording and rail in high speed railway straightness, foreign body intrusion detection, the inspection of Along Railway landslide of high compression ratio
Survey etc..Since the sound barrier used in certain locations is for lowering to surrounding enviroment noise jamming, while there is the function of protective fence
Can, but physically having larger difference with protective fence, therefore can also design according to different types of protective fence different
Damage condition detection algorithm.
(2) detection method of the automatic visual of the panorama sketch of the invention based on propulsion video, by the view of magnanimity
The panorama bitmap-format of frequency lightweight accordingly carries out lossless information extraction, not only reduces the storage and visit of video data form
Ask expense, and convert video to it is a kind of more suitable for manually inspecting and the form of computer analysis processing, solve by
Propulsion video is quickly converted to the undistorted of panorama sketch.It is saved greatly in addition, storing video information in the form of panorama sketch
The memory space of amount replaces propulsion video image as test object, it can be achieved that high-speed railway seals in the form of panorama sketch
The quick detection in closed loop border, meets actual demand.
(3) scene acquisition modes of the video capture of propulsion as a kind of movable type, since its visual field is broad, the depth of field
Far, space broad covered area has been widely used in mobile scene record and monitor task.The present invention is sealed for high-speed rail
Running environment state-detection is closed, a kind of video based on propulsion is proposed, utilizes the high-speed iron of computer vision technique
The automatic rapid detection method of road protective fence damage condition, to complete high-speed railway closed operation using high speed comprehensive detection train
Ambient condition detection provides a kind of new efficient means.
(4) the present invention is based on the virtual sampling channel models of the propulsion video of detection zone geometry, before overcoming
The static fuzzy spliced to sport video narrowband, and using the priori of region geometry structure, construct quick propulsion video
Panoramic mosaic algorithm, using railway protective column defect detection as application study background, propose accordingly based on panorama sketch from
Dynamicization visual detection algorithm.High-speed railway environment measuring is regular, normality a job, periodically obtains ambient video figure repeatedly
Picture.The information for obtaining abnormality is compared by feature detection or scene image.The present invention is mentioned for high-speed railway environment measuring
The new splicing technology of panorama drawing of one kind based on high-speed forward sport video is gone out.
(5) it the present invention is based on the virtual rectangle sampling channel Panoramagram montage method of propulsion video, realizes for the first time
Propulsion video panorama under camera optical axis and direction of motion uniform condition generates；Using virtual rectangle sampling channel side
Formula overcomes propulsion video panorama static distortion problem；Based on the complete of propulsion video virtual rectangle sampling channel
Scape figure generates, and does not have particular/special requirement for equipment, the video of general automobile data recorder record can be spliced into a variety of different use
Way panorama sketch.
Claims (6)
1. a kind of Panoramagram montage method based on virtual sampling channel model characterized by comprising
Obtain the propulsion video of record high-speed railway ambient condition；It is N's that frame number is extracted from the propulsion video
Sequence of video images；
According to the railway scene structure determined via end point, external sampling rectangle OR is set in every frame video imagem；It disappears
Coordinate of the point in image coordinate system is (x0,y0)T；The analytic expression of i-th line section is x+kiy+bi=0, wherein kiIt is i-th
The slope of line segment, biFor intercept；Weight wiIt is the length of i-th line section；N is total columns of pixel in protective fence panorama sketch；
Wherein,
According to the speed of train, internal sample rectangle IR is set in every frame video imagem；
It will be by the external sampling rectangle OR of every frame video imagemWith the internal sample rectangle IRmThe rectangular ring of composition is adopted
Sample annulus is divided into four strips mosaic region St,Sb,Sl,Sr；
By four strips mosaic regions S of every frame video imaget,Sb,Sl,SrBy image volume around, be corrected to rule square
Shape strips St',Sb',Sl',Sr'；
By 4 × the N number of corrected shape of rectangular ribbon St',Sb',Sl',Sr', split is carried out respectively according to respective sample plane,
Generate the panorama sketch of 4 planes of railway scene.
2. a kind of damage condition of the protective fence of the detection high-speed railway based on Panoramagram montage method described in claim 1
Method characterized by comprising
Step 1 obtains the propulsion video of record high-speed railway ambient condition；
The propulsion Video Quality Metric is railway based on the Panoramagram montage method of virtual sampling channel model by step 2
The left and right panorama sketch of running environment, as protective fence panorama sketch；
Step 3 is divided principle according to entropy is maximized, is determined the position of guardrail and background in the protective fence panorama sketch
, it indicates the position of protective fence, the position of background to be indicated at F (j)=0 at F (j)=1, obtains the two-value code of 0 and 1 composition；J is
The column serial number of protective fence panorama sketch；
Step 4 carries out run-length encoding to the two-value code；
Step 5 calculates the pixel wide D of the background area between adjacent two guardrail according to the run-length encodingcur；
Step 6, according to the pixel wide D of the background area between adjacent two guardrailcurBetween adjacent two guardrail
Normal pixel spacing d, judge that adjacent two guardrail with the presence or absence of defect, generates judging result；
Step 7 exports the judging result.
3. according to the method described in claim 2, it is characterized in that, the step 6 includes:
Work as DcurWhen greater than kd, then the judging result are as follows: there are defects for adjacent two guardrail；Wherein, k is experience adjusting
Parameter.
4. according to the method described in claim 2, it is characterized in that, the step 2 includes:
The sequence of video images that frame number is N is extracted from the propulsion video；
According to the railway scene structure determined via end point, external sampling rectangle OR is set in every frame video imagem；
According to the speed of train, internal sample rectangle IR is set in every frame video imagem；
By every frame video image by the external sampling rectangle ORmWith the internal sample rectangle IRmThe rectangular ring of composition is adopted
Sample annulus is divided into four strips mosaic region St,Sb,Sl,Sr；Wherein, St,Sb,Sl,SrRespectively sky, rail, left side shield
Column and right side guardrail region；
By four strips mosaic regions S of every frame video imaget,Sb,Sl,SrBy image volume around, be corrected to rule square
Shape strips St',Sb',Sl',Sr'；
By 2 × the N number of corrected shape of rectangular ribbon Sl',Sr', split is carried out respectively according to respective sample plane, generates iron
The left and right panorama sketch of road scene.
5. according to the method described in claim 2, it is characterized in that, the step 3 includes:
For each column j of protective fence panorama sketch, calculate gray average M (1, j), standard deviation V (1, j) and gradient mean value G (1,
j)；J=1 ..., N；N is total columns of pixel in protective fence panorama sketch；
Using F (j) maximum entropy, segmentation threshold (ε is calculated*,ξ*,η*)；F (j) is the vertical side extracted in the protective fence panorama sketch
The feature distribution of upward each column pixel；
According to the segmentation threshold (ε*,ξ*,η*), binarization segmentation is carried out to F (j), protective fence region is set to 1, background area
It is set to 0；
Wherein, M (1, j) is the gray average of each column pixel in protective fence panorama sketch；
V (1, j) is the gray variance of each column pixel in protective fence panorama sketch；G (1, j) is each column pixel in protective fence panorama sketch
Gradient mean value.
6. according to the method described in claim 5, it is characterized in that,
Wherein, p (i, j) is the gray value of pixel (i, j) in protective fence panorama sketch, th,bhThe respectively top and bottom of protective fence
Coordinate position of the portion in protective fence panorama sketch.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610312137.XA CN106023076B (en) | 2016-05-11 | 2016-05-11 | The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610312137.XA CN106023076B (en) | 2016-05-11 | 2016-05-11 | The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106023076A CN106023076A (en) | 2016-10-12 |
| CN106023076B true CN106023076B (en) | 2019-04-23 |
Family
ID=57099345
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610312137.XA Expired - Fee Related CN106023076B (en) | 2016-05-11 | 2016-05-11 | The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106023076B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106709892A (en) * | 2016-11-15 | 2017-05-24 | 昂纳自动化技术（深圳）有限公司 | Rapid region expansion algorithm and device of any structure element based on stroke coding |
| CN106851398A (en) * | 2017-03-10 | 2017-06-13 | 北京小鸟看看科技有限公司 | The storage method of panoramic video file, player method, storage device and player |
| CN107103578B (en) * | 2017-03-28 | 2020-12-11 | 北京交通大学 | Method for carrying out panorama stitching by utilizing forward vehicle-mounted video of high-speed railway |
| JP6903477B2 (en) | 2017-04-21 | 2021-07-14 | 株式会社東芝 | Orbit identification device and program |
| CN109842761B (en) * | 2019-02-01 | 2021-03-19 | 中国社会科学院大学 | Rapid high-definition panoramic imaging method and device for forward motion video |
| CN111241900B (en) * | 2019-04-12 | 2021-02-05 | 宁夏爱特云翔信息技术有限公司 | Traffic environment field maintenance method |
| CN112837341B (en) * | 2021-01-26 | 2022-05-03 | 石家庄铁道大学 | Self-adaptive time-space domain pedestrian appearance restoration method |
| CN117532624B (en) * | 2024-01-10 | 2024-03-26 | 南京东奇智能制造研究院有限公司 | Automatic positioning and aligning method and system for guardrail plate installation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101621634A (en) * | 2009-07-24 | 2010-01-06 | 北京工业大学 | Method for splicing large-scale video with separated dynamic foreground |
| CN103236160A (en) * | 2013-04-07 | 2013-08-07 | 水木路拓科技(北京)有限公司 | Road network traffic condition monitoring system based on video image processing technology |
| CN103985254A (en) * | 2014-05-29 | 2014-08-13 | 四川川大智胜软件股份有限公司 | Multi-view video fusion and traffic parameter collecting method for large-scale scene traffic monitoring |
| CN105046649A (en) * | 2015-06-30 | 2015-11-11 | 硅革科技（北京）有限公司 | Panorama stitching method for removing moving object in moving video |
-
2016
- 2016-05-11 CN CN201610312137.XA patent/CN106023076B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101621634A (en) * | 2009-07-24 | 2010-01-06 | 北京工业大学 | Method for splicing large-scale video with separated dynamic foreground |
| CN103236160A (en) * | 2013-04-07 | 2013-08-07 | 水木路拓科技(北京)有限公司 | Road network traffic condition monitoring system based on video image processing technology |
| CN103985254A (en) * | 2014-05-29 | 2014-08-13 | 四川川大智胜软件股份有限公司 | Multi-view video fusion and traffic parameter collecting method for large-scale scene traffic monitoring |
| CN105046649A (en) * | 2015-06-30 | 2015-11-11 | 硅革科技（北京）有限公司 | Panorama stitching method for removing moving object in moving video |
Non-Patent Citations (5)
| Title |
|---|
| Railroad online: acquiring and visualizing route panoramas of rail scenes;Shengchun Wang 等;《VISUAL COMPUTER》;20140930;第30卷(第9期);1045-1057 |
| Rendering Railway Scenes in Cyberspace Based on Route Panoramas;Shengchun Wang 等;《2013 International Conference on Cyberworlds》;20131023;12-19 |
| ROUTE PANORAMA ACQUISITION AND RENDERING FOR HIGH-SPEED RAILWAY MONITORING;Shengchun Wang 等;《ICME 2013》;20130731;1-6 |
| 基于列车前向运动视频的全景图拼接算法;高大龙 等;《山东大学学报(工学版)》;20131231;第43卷(第6期);1-6 |
| 移动车载视频的立体全景图生成方法;王胜春 等;《光学学报》;20141231;第34卷(第12期);1215005-1至1215005-9 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106023076A (en) | 2016-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106023076B (en) | The method of the damage condition of the protective fence of the joining method and detection high-speed railway of panorama sketch | |
| CN106910203B (en) | The quick determination method of moving target in a kind of video surveillance | |
| CN108491758B (en) | Track detection method and robot | |
| CN106960179B (en) | Rail line Environmental security intelligent monitoring method and device | |
| CN104134222A (en) | Traffic flow monitoring image detecting and tracking system and method based on multi-feature fusion | |
| US11335036B2 (en) | Image synthesizing system and image synthesizing method | |
| CN113011252B (en) | Rail foreign matter intrusion detection system and method | |
| CN110255318B (en) | Method for detecting idle articles in elevator car based on image semantic segmentation | |
| CN113298059A (en) | Pantograph foreign matter intrusion detection method, device, computer equipment, system and storage medium | |
| CN113034439B (en) | High-speed railway sound barrier defect detection method and device | |
| US20200034626A1 (en) | Method for recognizing an object of a mobile unit | |
| CN111561967A (en) | Real-time online detection method and system for pantograph-catenary operation state | |
| CN104570764A (en) | Verification platform for airborne target indicating system | |
| CN115600124A (en) | Subway tunnel inspection system and inspection method | |
| CN116824859A (en) | Intelligent traffic big data analysis system based on Internet of things | |
| CN111598855B (en) | 2C equipment high-speed rail contact net dropper defect detection method based on deep learning and transfer learning | |
| CN111723704A (en) | Raspberry pie-based van body door opening monitoring method | |
| CN107169969B (en) | Highway dangerous rock collapse deposit size measurement and alarm system based on FPGA | |
| Wu et al. | Vehicle Classification and Counting System Using YOLO Object Detection Technology. | |
| Panchenko et al. | Improvement of the accuracy of determining movement parameters of cuts on classification humps by methods of video analysis | |
| CN112508893A (en) | Machine vision-based method and system for detecting tiny foreign matters between two railway tracks | |
| Jiang et al. | Railway panorama: A fast inspection method for high-speed railway infrastructure monitoring | |
| CN113362330B (en) | Pantograph cavel real-time detection method, device, computer equipment and storage medium | |
| Wolf et al. | Asset Detection in Railroad Environments using Deep Learning-based Scanline Analysis. | |
| Nichoha et al. | Development of modern methods and directions of rapid diagnostics of railway tracks defects by television methods |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190423 Termination date: 20200511 |