CN107153832A - A kind of high ferro contact net equipotential line releases detection method and system - Google Patents
A kind of high ferro contact net equipotential line releases detection method and system Download PDFInfo
- Publication number
- CN107153832A CN107153832A CN201610121292.3A CN201610121292A CN107153832A CN 107153832 A CN107153832 A CN 107153832A CN 201610121292 A CN201610121292 A CN 201610121292A CN 107153832 A CN107153832 A CN 107153832A
- Authority
- CN
- China
- Prior art keywords
- image
- region
- mtd
- mrow
- equipotential line
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- 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
- G06V10/457—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 analysing connectivity, e.g. edge linking, connected component analysis or slices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
Detection method and system are released the invention discloses a kind of high ferro contact net equipotential line, is related to electrified high ferro contact net safety testing field.This method includes：Gather the region of interest area image of each pillar of high ferro contact net in real time by image capture device；The region of interest area image includes locator bearing, locator and equipotential line；Region of interest area image is subjected to binary conversion treatment and morphological dilations processing, the image after expansion process is obtained；Connected component labeling is carried out to the image after expansion process, judges whether equipotential line releases according to the feature of connected region in image.The equipotential line provided using the present invention releases detection method and system, the quick detection to equipotential line state can be quickly finished, judge whether equipotential line releases, improve the efficiency and accuracy of equipotential line fault detect, to find the failure in high ferro contact net as early as possible, guarantee has been safely provided for high ferro contact net, potential safety hazard can be effectively reduced.
Description
Technical field
The present invention relates to electrified high ferro contact net safety testing field, and in particular to a kind of high ferro contact net
Equipotential line releases detection method and system.
Background technology
With the further development and its extensive use of high-speed railway transportation technology, high ferro contact net as to
Locomotive provides the critical facility of electric energy, and whether it also increasingly attracts attention in safe work state.It is real
In border, high ferro contact net failure it is main by parts it is loose, de-, lack, split etc. and to cause, equipotential line is
The important component of positioner is supported in contact net, whether in good condition its is directly determines electric power machine
The normal power supply of car.And for equipotential line part major defect to be loose and de-, at present, high ferro contact net
Detection pattern is the artificial inspection of reaching the standard grade in skylight operation, but because circuit local environment is complicated, it is artificial to see
The dangerous high and unequal factor of staff quality is examined, and is possible to occur the situation of missing inspection, so as to make
Into potential safety hazard.Therefore, how quick and precisely to detect high ferro contact net medium potential wire state be one urgently
To be solved the problem of.
The content of the invention
For defect present in prior art, it is an object of the invention to provide a kind of high ferro contact net etc.
Equipotential line releases detection method and system, to realize the real-time detection of equipotential line state.
To achieve the above object, the technical solution adopted by the present invention is：
A kind of high ferro contact net equipotential line releases detection method, comprises the following steps：
(1) the area-of-interest figure of each pillar of high ferro contact net is gathered in real time by image capture device
Picture；The region of interest area image includes locator bearing, locator and equipotential line；
(2) the region of interest area image is subjected to binary conversion treatment, obtains binary image；
(3) binary image is carried out by morphological dilations processing according to default circular configuration element,
Obtain the image after expansion process；
(4) image after expansion process is subjected to connected component labeling, according to the spy of connected region in image
Levy and judge whether equipotential line releases；The feature of the connected region includes area and the company of connected region
The quantity in logical region；Judge that the mode whether equipotential line releases is：
4.1) candidate's connected region in screening image；Candidate's connected region refers to connected region area
More than the connected region area defined of setting area threshold；
4.2) area according to where the feature of locator bearing determines locator bearing in candidate's connected region
Domain；The feature of the locator bearing includes the area and position feature of locator；
4.3) connected region of the area in the first setting areal extent in the region where locator bearing is judged
Whether the quantity in domain is more than 1, if so, then judging that equipotential line is not released, if it is not, then entering next
Step；
4.4) by the image inverse of candidate's connected region, the area where locator bearing after inverse is judged
Connected region of the area in the second setting areal extent whether there is in domain, if so, then judging equipotential
Line is not released, if it is not, then judging that equipotential line is released.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step (2)
In, before region of interest area image progress binary conversion treatment, in addition to：
A. the template image of locator bearing is made, region of interest area image is matched using template image,
Determine the region where locator bearing in region of interest area image；
B. the region where the locator bearing that basis is determined is by region of interest area image according to pre-set image
Size is cut, using the image after cutting as new region of interest area image, to new region of interest
Area image carries out binary conversion treatment；The new region of interest area image includes locator bearing and waits electricity
Bit line.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step a
In, region of interest area image is matched using template image, locator in region of interest area image is determined
Region where bearing, including：
The matching degree ModelScore of every piece of region and template image in region of interest area image is calculated, according to
The number N of locator bearing in region of interest area image, by the N number of area of matching degree ModelScore highests
Domain is defined as the region where locator；1≤N≤2.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step b
In, the region according to where the locator bearing determined is big according to pre-set image by region of interest area image
It is small to be cut, including：
B1. matrixing is normalized in region of interest area image, makes locator in the image after conversion
The centre coordinate in the region where bearing is overlapped with the centre coordinate of template image；The matrixing includes
Translation transformation, scale transformation and rotation transformation；
B2. centered on the centre coordinate of template image, according to pre-set image size, after matrixing
Image to be matched cut.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step b1
In, region of interest area image is normalized to the normalization matrix HomMat2DGlobal of matrixing
Formula be：
HomMat2DGlobal=
HomMat2DTranSlate*HomMat2DScale*HomMat2DRotate
Wherein, HomMat2DTranSlate represents translation transformation matrix, and HomMat2DScale represents contracting
Transformation matrix is put, HomMat2DRotate represents rotational transformation matrix；
The translation transformation matrix calculation is：
Wherein, ModelRow, ModelColumn represent respectively the center of template image row coordinate and
Row coordinate, Row, Column represent respectively the center in the region where locator bearing row coordinate and
Row coordinate；
The scale transformation matrix is：
Wherein, when ModelScale represents image normalization to be matched, region of interest area image is relative to mould
The zoom factor of plate image；
The rotational transformation matrix is：
Wherein, when phi represents image normalization to be matched, region of interest area image is relative to template image
The anglec of rotation.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step (2)
In, region of interest area image is carried out by binary conversion treatment using local threshold algorithm.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step (2)
In, it is described that region of interest area image is subjected to binary conversion treatment, including：
2.1) region of interest area image is divided into by several regions according to setting regions size；
2.2) the segmentation threshold threshold in each region is set；
2.3) binary conversion treatment, binary conversion treatment are carried out to each region according to the segmentation threshold in each region
Formula be：
Wherein, f (x, y) is the image intensity value before binary conversion treatment, and g (x, y) is after binary conversion treatment
Image intensity value.
Further, a kind of high ferro contact net equipotential line as described above releases detection method, step (1)
In, in addition to：The region of interest area image is pre-processed；The pretreatment includes image enhaucament
And denoising；Described image strengthens the formula handled：
G '=g × Mult+Mult × GMin
Mult=255/ (GMax-GMin)
Wherein, g is the gray value of pixel in enhancing before processing image, and g ' is in image after enhancing processing
Pixel gray value；Mult is linear transform coefficient；GMax and GMin represent to strengthen before processing respectively
Maximum gradation value and minimum gradation value in image.
Detecting system is released present invention also offers a kind of high ferro contact net equipotential line, including：
Area-of-interest image capture module, for gathering high ferro contact net in real time by image capture device
Each pillar region of interest area image；The region of interest area image includes locator bearing, determined
Position device and equipotential line；
Image binaryzation module, for the region of interest area image to be carried out into binary conversion treatment, obtains two
Value image；
Image expansion processing module, for being entered the binary image according to default circular configuration element
The processing of row morphological dilations, obtains the image after expansion process；
Equipotential line detection module, for carrying out connected component labeling to the image after expansion process, according to
The feature of connected region judges whether equipotential line releases in image；The feature of the connected region includes
The area of connected region and the quantity of connected region；The module includes：
Candidate region screening unit, for screening candidate's connected region in image；The candidate connected region
Domain refers to that connected region area is more than the connected region area defined of setting area threshold；
Locator standoff region positioning unit, for determining that candidate connects according to the feature of locator bearing
Region in region where locator bearing；The area of the feature of the locator bearing including locator and
Position feature；
First judging unit, for judging that area is in setting areal extent in the region where locator bearing
Whether the quantity of interior connected region is more than 1, if so, then judging that equipotential line is not released, if it is not, then
Then enter the second judging unit；
Second judging unit, for by the image inverse of candidate's connected region, judging to position after inverse
Connected region of the area in the second setting areal extent whether there is in region where device bearing, if
It is then to judge that equipotential line is not released, if it is not, then judging that equipotential line is released.
Further, a kind of high ferro contact net equipotential line as described above releases detecting system, and the system is also
Including：
Image cropping module is right for before the region of interest area image is carried out into binary conversion treatment
Region of interest area image is cut；The module includes：
Locator bearing determining unit, the template image for making locator bearing, by area-of-interest
Image is matched using template image, determines the region where locator bearing in region of interest area image；
Image cropping unit is big according to pre-set image for the region where the locator bearing determined
It is small that region of interest area image is cut, using the image after cutting as new region of interest area image,
Binary conversion treatment is carried out to new region of interest area image；The new region of interest area image includes fixed
Position device bearing and equipotential line.
The beneficial effects of the present invention are：Method and system provided by the present invention, can be quickly finished
To the quick detection of equipotential line state, judge whether equipotential line releases, improve equipotential line event
Hinder the efficiency and accuracy of detection, be high ferro contact net to find the failure in high ferro contact net as early as possible
Guarantee has been safely provided, potential safety hazard can be effectively reduced.
Brief description of the drawings
Fig. 1 releases detection method for a kind of high ferro contact net equipotential line in the specific embodiment of the invention
Flow chart；
Fig. 2 is the structural representation of positioning support device medium potential line part in high ferro contact net；
Fig. 3 releases detecting system for a kind of high ferro contact net equipotential line in the specific embodiment of the invention
Structured flowchart；
Fig. 4 be embodiment in one cut after new region of interest area image schematic diagram；
Fig. 5 is the image after Fig. 4 a form of expansion process；
Fig. 6 is the change image after Fig. 4 another form of expansion process；
Fig. 7 is the schematic diagram of locator bearing region in Fig. 5；
Fig. 8 is by the partial schematic diagram after Fig. 6 inverses.
Embodiment
With reference to Figure of description, the present invention is described in further detail with embodiment.
Fig. 1 shows that a kind of high ferro contact net equipotential line provided in the specific embodiment of the invention is released
The flow chart of detection method, as can be seen from Figure, this method may comprise steps of：
Step S100：The sense for gathering each pillar of high ferro contact net in real time by image capture device is emerging
Interesting area image；
Fig. 2 shows the schematic diagram in positioning support device medium potential line part, figure in high ferro contact net
In show equipotential line 1, locator bearing 2 and locator 3 etc..In order to judge high ferro contact net
Whether medium potential line releases, it is necessary first to gather in the image of equipotential portion, present embodiment, leads to
The region of interest area image that image capture device gathers each pillar of high ferro contact net in real time is crossed, it is interested
At least include locator bearing, the part-structure and equipotential line of locator in area image.
In actual applications, in order to improve the speed of Data Management Analysis, enter to region of interest area image
Before row processing, processing can be zoomed in and out to the original image of collection first, the part of scaling can root
Set according to being actually needed.
, can be by institute in order to more protrude the feature of locator bearing and equipotential line when handling image
Region of interest area image is stated to be pre-processed；The pretreatment includes image enhaucament and denoising.This reality
Apply in mode, the formula of described image enhancing processing is：
G '=g × Mult+Mult × GMin
Mult=255/ (GMax-GMin)
Wherein, g is the gray value of pixel in enhancing before processing image, and g ' is in image after enhancing processing
The gray value of pixel；Mult is linear transform coefficient；GMax and GMin represent that enhancing is handled respectively
Maximum gradation value and minimum gradation value in preceding image.
Complete after image enhancement processing, can be again using Gaussian filter to various noises present in image
(noise that the noise and image enhancement processes existed during collection is produced) carries out denoising smooth processing.
Step S200：The region of interest area image is subjected to binary conversion treatment, binary picture is obtained
Picture；
In order that prospect (locator bearing, locator, equipotential line) in region of interest area image with
Background separation, binary conversion treatment, the binary picture after being handled are carried out by the region of interest area image
Picture.
In present embodiment, in order to further improve image processing efficiency, enter by region of interest area image
Before row binary conversion treatment, in addition to the step of cut to region of interest area image, pass through the step
Cut out and go locator bearing in image, equipotential line, the portion locators structure being directly connected to equipotential line
Outside other regions.In present embodiment, the concrete mode cut to region of interest area image
For：
A. the template image of locator bearing is made, region of interest area image is matched using template image,
The region where locator bearing in region of interest area image is determined, specifically：It is interested by calculating
The matching degree ModelScore of every piece of region and template image in area image, according to region of interest area image
The number N of middle locator bearing, locator is defined as by the N number of region of matching degree ModelScore highests
The region at place；1≤N≤2.Due in actual job, the support of each pillar of high ferro contact net
Have plenty of a locator bearing in positioner, have plenty of two locator bearings, therefore, root
Factually one or two matching degree highest region is defined as where locator bearing by the application scenarios on border
Region.The specific of corresponding structure in region of interest area image is matched according to the template image of a structure
Mode is prior art, is not described in detail herein.
In matching process, the center every trade coordinate ModelRow and row coordinate of logging template image
ModelColumn, region of interest area image need the angle ModelAngle and region of interest rotated
Zoom factor ModelScale of the area image relative to template image.In present embodiment, the origin of coordinates is
The upper left corner of image.
B. the region and locator bearing where the locator bearing that basis is determined are in area-of-interest figure
Position as in, cuts according to pre-set image size to region of interest area image, with the figure after cutting
As new region of interest area image, binary conversion treatment is carried out to new region of interest area image；It is described
New region of interest area image includes locator bearing and equipotential line.
In present embodiment, the region and locator bearing according to where locator bearing are in region of interest
Position in area image, according to pre-set image size (such as image size is 1000*1000) to interested
The concrete mode that area image is cut is：
B1. matrixing is normalized in region of interest area image, makes locator in the image after conversion
The centre coordinate in the region where bearing is overlapped with the centre coordinate of template image；The matrixing includes
Translation transformation, scale transformation and rotation transformation；
B2. centered on the centre coordinate of template image, according to pre-set image size, after matrixing
Image to be matched cut.
Wherein, in step b1, region of interest area image is normalized to the normalized moments of matrixing
Battle array HomMat2DGlobal formula be：
HomMat2DGlobal=
HomMat2DTranSlate*HomMat2DScale*HomMat2DRotate
Wherein, HomMat2DTranSlate represents translation transformation matrix, and HomMat2DScale represents contracting
Transformation matrix is put, HomMat2DRotate represents rotational transformation matrix；
The translation transformation matrix calculation is：
Wherein, ModelRow, ModelColumn represent the row coordinate and row at the center of template image respectively
Coordinate, Row, Column represent the row coordinate and row at the center in the region where locator bearing respectively
Coordinate；
The scale transformation matrix is：
Wherein, when ModelScale represents image normalization to be matched, region of interest area image is relative to mould
The zoom factor of plate image；
The rotational transformation matrix is：
Wherein, when phi represents image normalization to be matched, region of interest area image is relative to template image
The anglec of rotation (locator bearing region), i.e. phi=ModelAngle.
Because in actual treatment, pending region of interest area image is numerous, by present embodiment
The normalized of image is realized in the above-mentioned matrixing provided, pending image is reached unification
Picture format requirement.
In order to more protrude in the difference of Each part in image, present embodiment, using local threshold
Region of interest area image is carried out binary conversion treatment by algorithm, and specific processing mode is：
1) region of interest area image is divided into by several regions according to setting regions size.
2) the segmentation threshold threshold in each region is set.
3) binary conversion treatment is carried out to each region according to the segmentation threshold in each region, binary conversion treatment
Formula is：
Wherein, f (x, y) is the image intensity value before binary conversion treatment, and g (x, y) is after binary conversion treatment
Image intensity value.
Step S300：Binary image is subjected to morphological dilations processing, the figure after expansion process is obtained
Picture；
Step S400：Connected component labeling is carried out to binary image, according to connected region in image
Feature judges whether equipotential line releases.
After the binary conversion treatment for completing image, in order that directly being contacted with locator bearing, equipotential line
Part can be also entered in follow-up analysis deterministic process, to improve the degree of accuracy of subsequent analysis, according to
The binary image is carried out morphological dilations processing by default circular configuration element, obtains expansion process
Image afterwards.Wherein, the radius of circular configuration element is set according to actual needs, generally experience
When value, such as image size are 1000*1000, the radius of circular configuration element could be arranged to 5.This reality
Apply in mode, the unit of picture size is number of pixels, and the number of pixels that 1000*1000 represents total is
1000000.The radius of the circular configuration is also 5 pixels for 5 fingers.
After the expansion process for completing image, the connected region in the image after expansion process is marked,
And the judgement of equipotential line state is completed according to the feature of each connected region.In present embodiment, the company
The feature in logical region includes the area of connected region and the quantity of connected region, whether judges equipotential line
The concrete mode released is：
1) candidate's connected region in screening image；Candidate's connected region refers to that connected region area is big
In the connected region area defined of setting area threshold；To area-of-interest image processing process
In, there may be some reasons in the image after expansion process in the enhancing of image, denoising and two
Value etc. is handled and small connected region, and the connection that not real image is truly present of these connected regions
Region, therefore, in order to improve the precision of detection, first by setting the setting area threshold, by two
Area is screened out less than the connected region of the threshold value in value image, the area that the connected region after screening is surrounded
Domain is referred to as candidate's connected region；
2) area according to where the feature of locator bearing determines locator bearing in candidate's connected region
Domain；The feature of the locator bearing includes the area and position feature of locator；
Because other structures are present in locator bearing and its position and image in region of interest area image
It is obvious different, therefore, the connected region of locator bearing region in the image after above-mentioned expansion process
The feature in domain connected region corresponding with other structures also can be significantly different, therefore, it can according at expansion
The feature (area, position) of each connected region determines the region where locator bearing in image after reason,
Specifically, connected domain area and connected domain position meet preset area scope and pre- in candidate's connected region
If the connected domain of position is locator bearing region.As shown in figure 5, for by shown in Fig. 4
Determining in image of the image after binary conversion treatment, the Fig. 5 determined according to the feature of locator bearing
Position device bearing region is the shown rectangular area in Fig. 7.
3) connected region of the area in the first setting areal extent in the region where locator bearing is judged
Quantity whether be more than 1, if so, then judge that equipotential line is not released, if it is not, then into next step；
4) by the image inverse of candidate's connected region, the region where locator bearing after inverse is judged
Connected region of the middle area in the second setting areal extent whether there is, if so, then judging equipotential line
Do not release, if it is not, then judging that equipotential line is released.
In actual process, due to the region of interest area image gathered every time by image capture device
The gray value of middle prospect and background parts all can be different, therefore, though in step s 200 using
Identical binary conversion treatment mode, the binary conversion treatment structure of identical image is also possible to difference, and this just leads
Cause to need the image of detection different after follow-up expansion processing, be to enter Fig. 4 as shown in Figure 5 and Figure 6
Image after row binaryzation and expansion process, and in two images there is very big difference in the feature of connected region
It is different.In order to improve the accuracy rate of judged result, it is to avoid cause misjudgment because of the difference of binary image
The problem of, by above-mentioned steps 2 in present embodiment) and step 3) realize different binaryzation results
Judge, image after binaryzation as shown in Figure 5 and expansion, image medium potential line and locator bearing
Color on the contrary, equipotential line by locator standoff region divide for two connected regions, i.e. locator
Meet the quantity of the first connected region for setting areal extent in bearing region as 2, it can be determined that go out
Equipotential line is not released, and in the image shown in Fig. 6 locator bearing region connected region quantity
For 1, at this moment again by the way that the image in Fig. 6 is carried out after inverse, inverse into image as shown in figure 8, by anti-
If the image after color, which can be seen that in the image after equipotential line presence, inverse, can deposit institute in fig. 8
The polygonal region (oblique line in figure marks region) shown, and the part be by locator bearing, etc. electricity
The region that bit line and locator are surrounded, its area is within a fixed range, to therefore, it can basis
It whether there is in image after inverse and meet the connected region of the second setting areal extent to judge equipotential line
Whether release.Wherein, it is described first setting areal extent and second setting areal extent be empirical value.
In present embodiment, the area of connected region refers to the number of pixels contained in connected region.
Detection method is released by the above-mentioned equipotential line provided in present embodiment, equipotential can be realized
The quick real-time detection that line is released, is to find that contact net failure provides warning information and provided the foundation as early as possible, protects
Demonstrate,prove contact net and be in good working state, in favor of the safe operation of electric railway.
Corresponding to the method shown in Fig. 1, a kind of electricity such as high ferro contact net are additionally provided in present embodiment
Bit line releases detecting system, as shown in figure 3, the system includes area-of-interest image capture module
100, image cropping module 200 and is waited image binaryzation module 300, image expansion processing module 400
Equipotential line detection module 500.
Area-of-interest image capture module 100, connects for gathering high ferro in real time by image capture device
The region of interest area image for each pillar touched net；The region of interest area image includes locator branch
Seat, locator and equipotential line；
Image cropping module 200, for by the region of interest area image carry out binary conversion treatment it
Before, to region of interest area image is cut；The module includes：
Locator bearing determining unit, the template image for making locator bearing, by area-of-interest
Image is matched using template image, determines the region where locator bearing in region of interest area image；
Image cropping unit, for the region according to where the locator bearing determined and locator branch
Position of the seat in region of interest area image, cuts out according to pre-set image size to region of interest area image
Cut, using the image after cutting as new region of interest area image, new region of interest area image is carried out
Binary conversion treatment；The new region of interest area image includes locator bearing and equipotential line.
Image binaryzation module 300, for the region of interest area image to be carried out into binary conversion treatment, is obtained
To binary image；
Image expansion processing module 400, for according to default circular configuration element by the binary picture
As carrying out morphological dilations processing, the image after expansion process is obtained；
Equipotential line detection module 500, for carrying out connected component labeling to the image after expansion process,
Judge whether equipotential line releases according to the feature of connected region in image；The feature of the connected region
The quantity of area and connected region including connected region；The module includes：
Candidate region screening unit, for screening candidate's connected region in image；The candidate connected region
Domain refers to that connected region area is more than the connected region area defined of setting area threshold；
Locator standoff region positioning unit, for determining that candidate connects according to the feature of locator bearing
Region in region where locator bearing；The area of the feature of the locator bearing including locator and
Position feature；
First judging unit, for judging that area is in setting areal extent in the region where locator bearing
Whether the quantity of interior connected region is more than 1, if so, then judging that equipotential line is not released, if it is not, then
Then enter the second judging unit；
Second judging unit, for by the image inverse of candidate's connected region, judging to position after inverse
Connected region of the area in the second setting areal extent whether there is in region where device bearing, if
It is then to judge that equipotential line is not released, if it is not, then judging that equipotential line is released.
Obviously, those skilled in the art can carry out various changes and modification without departing from this to the present invention
The spirit and scope of invention.So, if these modifications and variations of the present invention belong to right of the present invention and wanted
Ask and its equivalent technology within the scope of, then the present invention be also intended to comprising these change and modification including.
Claims (10)
1. a kind of high ferro contact net equipotential line releases detection method, comprise the following steps：
(1) the area-of-interest figure of each pillar of high ferro contact net is gathered in real time by image capture device
Picture；The region of interest area image includes locator bearing, locator and equipotential line；
(2) the region of interest area image is subjected to binary conversion treatment, obtains binary image；
(3) binary image is carried out by morphological dilations processing according to default circular configuration element,
Obtain the image after expansion process；
(4) image after expansion process is subjected to connected component labeling, according to the spy of connected region in image
Levy and judge whether equipotential line releases；The feature of the connected region includes area and the company of connected region
The quantity in logical region；Judge that the mode whether equipotential line releases is：
4.1) candidate's connected region in screening image；Candidate's connected region refers to connected region area
More than the connected region area defined of setting area threshold；
4.2) area according to where the feature of locator bearing determines locator bearing in candidate's connected region
Domain；The feature of the locator bearing includes the area and position feature of locator；
4.3) connected region of the area in the first setting areal extent in the region where locator bearing is judged
Whether the quantity in domain is more than 1, if so, then judging that equipotential line is not released, if it is not, then entering next
Step；
4.4) by the image inverse of candidate's connected region, the area where locator bearing after inverse is judged
Connected region of the area in the second setting areal extent whether there is in domain, if so, then judging equipotential
Line is not released, if it is not, then judging that equipotential line is released.
2. a kind of high ferro contact net equipotential line according to claim 1 releases detection method, its
It is characterised by：In step (2), before region of interest area image progress binary conversion treatment, also wrap
Include：
A. the template image of locator bearing is made, region of interest area image is matched using template image,
Determine the region where locator bearing in region of interest area image；
B. the region where the locator bearing that basis is determined is by region of interest area image according to pre-set image
Size is cut, using the image after cutting as new region of interest area image, to new region of interest
Area image carries out binary conversion treatment；The new region of interest area image includes locator bearing and waits electricity
Bit line.
3. a kind of high ferro contact net equipotential line according to claim 2 releases detection method, its
It is characterised by：In step a, region of interest area image is matched using template image, determined interested
Region in area image where locator bearing, including：
The matching degree ModelScore of every piece of region and template image in region of interest area image is calculated, according to
The number N of locator bearing in region of interest area image, by the N number of area of matching degree ModelScore highests
Domain is defined as the region where locator；1≤N≤2.
4. a kind of high ferro contact net equipotential line according to claim 2 releases detection method, its
It is characterised by：In step b, region according to where the locator bearing determined is by area-of-interest figure
As being cut according to pre-set image size, including：
B1. matrixing is normalized in region of interest area image, makes locator in the image after conversion
The centre coordinate in the region where bearing is overlapped with the centre coordinate of template image；The matrixing includes
Translation transformation, scale transformation and rotation transformation；
B2. centered on the centre coordinate of template image, according to pre-set image size, after matrixing
Image to be matched cut.
5. a kind of high ferro contact net equipotential line according to claim 4 releases detection method, its
It is characterised by：In step b1, region of interest area image is normalized to the normalized moments of matrixing
Battle array HomMat2DGlobal formula be：
HomMat2DGlobal=
HomMat2DTranSlate*HomMat2DScale*HomMat2DRotate
Wherein, HomMat2DTranSlate represents translation transformation matrix, and HomMat2DScale represents contracting
Transformation matrix is put, HomMat2DRotate represents rotational transformation matrix；
The translation transformation matrix calculation is：
<mrow>
<mi>H</mi>
<mi>o</mi>
<mi>m</mi>
<mi>M</mi>
<mi>a</mi>
<mi>t</mi>
<mn>2</mn>
<mi>D</mi>
<mi>T</mi>
<mi>r</mi>
<mi>a</mi>
<mi>n</mi>
<mi>S</mi>
<mi>l</mi>
<mi>a</mi>
<mi>t</mi>
<mi>e</mi>
<mo>=</mo>
<mfenced open = "[" close = "]">
<mtable>
<mtr>
<mtd>
<mn>1</mn>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mrow>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mi>l</mi>
<mi>R</mi>
<mi>o</mi>
<mi>w</mi>
<mo>-</mo>
<mi>R</mi>
<mi>o</mi>
<mi>w</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>1</mn>
</mtd>
<mtd>
<mrow>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mi>l</mi>
<mi>C</mi>
<mi>o</mi>
<mi>l</mi>
<mi>u</mi>
<mi>m</mi>
<mi>n</mi>
<mo>-</mo>
<mi>C</mi>
<mi>o</mi>
<mi>l</mi>
<mi>u</mi>
<mi>m</mi>
<mi>n</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>1</mn>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, ModelRow, ModelColumn represent respectively the center of template image row coordinate and
Row coordinate, Row, Column represent respectively the center in the region where locator bearing row coordinate and
Row coordinate；
The scale transformation matrix is：
<mrow>
<mi>H</mi>
<mi>o</mi>
<mi>m</mi>
<mi>M</mi>
<mi>a</mi>
<mi>t</mi>
<mn>2</mn>
<mi>D</mi>
<mi>S</mi>
<mi>c</mi>
<mi>a</mi>
<mi>l</mi>
<mi>e</mi>
<mo>=</mo>
<mfenced open = "[" close = "]">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mi>l</mi>
<mi>S</mi>
<mi>c</mi>
<mi>a</mi>
<mi>l</mi>
<mi>e</mi>
</mrow>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mrow>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mi>l</mi>
<mi>S</mi>
<mi>c</mi>
<mi>a</mi>
<mi>l</mi>
<mi>e</mi>
</mrow>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>1</mn>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, when ModelScale represents image normalization to be matched, region of interest area image is relative to mould
The zoom factor of plate image；
The rotational transformation matrix is：
<mrow>
<mi>H</mi>
<mi>o</mi>
<mi>m</mi>
<mi>M</mi>
<mi>a</mi>
<mi>t</mi>
<mn>2</mn>
<mi>D</mi>
<mi>R</mi>
<mi>o</mi>
<mi>t</mi>
<mi>a</mi>
<mi>t</mi>
<mi>e</mi>
<mo>=</mo>
<mfenced open = "[" close = "]">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>cos</mi>
<mrow>
<mo>(</mo>
<mi>p</mi>
<mi>h</mi>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>sin</mi>
<mrow>
<mo>(</mo>
<mi>p</mi>
<mi>h</mi>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mo>-</mo>
<mi>sin</mi>
<mrow>
<mo>(</mo>
<mi>p</mi>
<mi>h</mi>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>cos</mi>
<mrow>
<mo>(</mo>
<mi>p</mi>
<mi>h</mi>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mn>1</mn>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, when phi represents image normalization to be matched, region of interest area image is relative to template image
The anglec of rotation.
6. a kind of high ferro contact net equipotential line according to one of claim 1 to 5 releases detection side
Method, it is characterised in that：In step (2), region of interest area image is carried out two using local threshold algorithm
Value is handled.
7. a kind of high ferro contact net equipotential line according to claim 6 releases detection method, its
It is characterised by：It is described that region of interest area image is subjected to binary conversion treatment in step (2), including：
2.1) region of interest area image is divided into by several regions according to setting regions size；
2.2) the segmentation threshold threshold in each region is set；
2.3) binary conversion treatment, binary conversion treatment are carried out to each region according to the segmentation threshold in each region
Formula be：
<mrow>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mn>1</mn>
</mtd>
<mtd>
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>&GreaterEqual;</mo>
<mi>t</mi>
<mi>h</mi>
<mi>r</mi>
<mi>e</mi>
<mi>s</mi>
<mi>h</mi>
<mi>o</mi>
<mi>l</mi>
<mi>d</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo><</mo>
<mi>t</mi>
<mi>h</mi>
<mi>r</mi>
<mi>e</mi>
<mi>s</mi>
<mi>h</mi>
<mi>o</mi>
<mi>l</mi>
<mi>d</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, f (x, y) is the image intensity value before binary conversion treatment, and g (x, y) is after binary conversion treatment
Image intensity value.
8. a kind of high ferro contact net equipotential line according to one of claim 1 to 5 releases detection side
Method, it is characterised in that：In step (1), in addition to：The region of interest area image is located in advance
Reason；The pretreatment includes image enhaucament and denoising；Described image strengthens the formula handled：
G '=g × Mult+Mult × GMin
Mult=255/ (GMax-GMin)
Wherein, g is the gray value of pixel in enhancing before processing image, and g ' is in image after enhancing processing
The gray value of pixel；Mult is linear transform coefficient；GMax and GMin represent that enhancing is handled respectively
Maximum gradation value and minimum gradation value in preceding image.
9. a kind of high ferro contact net equipotential line releases detecting system, including：
Area-of-interest image capture module, for gathering high ferro contact net in real time by image capture device
Each pillar region of interest area image；The region of interest area image includes locator bearing, determined
Position device and equipotential line；
Image binaryzation module, for the region of interest area image to be carried out into binary conversion treatment, obtains two
Value image；
Image expansion processing module, for being entered the binary image according to default circular configuration element
The processing of row morphological dilations, obtains the image after expansion process；
Equipotential line detection module, for carrying out connected component labeling to the image after expansion process, according to
The feature of connected region judges whether equipotential line releases in image；The feature of the connected region includes
The area of connected region and the quantity of connected region；The module includes：
Candidate region screening unit, for screening candidate's connected region in image；The candidate connected region
Domain refers to that connected region area is more than the connected region area defined of setting area threshold；
Locator standoff region positioning unit, for determining that candidate connects according to the feature of locator bearing
Region in region where locator bearing；The area of the feature of the locator bearing including locator and
Position feature；
First judging unit, for judging that area is in setting areal extent in the region where locator bearing
Whether the quantity of interior connected region is more than 1, if so, then judging that equipotential line is not released, if it is not, then
Then enter the second judging unit；
Second judging unit, for by the image inverse of candidate's connected region, judging to position after inverse
Connected region of the area in the second setting areal extent whether there is in region where device bearing, if
It is then to judge that equipotential line is not released, if it is not, then judging that equipotential line is released.
10. a kind of high ferro contact net equipotential line according to claim 9 releases detecting system, its
It is characterised by：The system also includes：
Image cropping module is right for before the region of interest area image is carried out into binary conversion treatment
Region of interest area image is cut；The module includes：
Locator bearing determining unit, the template image for making locator bearing, by area-of-interest
Image is matched using template image, determines the region where locator bearing in region of interest area image；
Image cropping unit is big according to pre-set image for the region where the locator bearing determined
It is small that region of interest area image is cut, using the image after cutting as new region of interest area image,
Binary conversion treatment is carried out to new region of interest area image；The new region of interest area image includes fixed
Position device bearing and equipotential line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610121292.3A CN107153832B (en) | 2016-03-03 | 2016-03-03 | A kind of high iron catenary equipotential line loosens detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610121292.3A CN107153832B (en) | 2016-03-03 | 2016-03-03 | A kind of high iron catenary equipotential line loosens detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107153832A true CN107153832A (en) | 2017-09-12 |
| CN107153832B CN107153832B (en) | 2019-08-30 |
Family
ID=59791919
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610121292.3A Active CN107153832B (en) | 2016-03-03 | 2016-03-03 | A kind of high iron catenary equipotential line loosens detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107153832B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108189626A (en) * | 2017-12-28 | 2018-06-22 | 深圳市灵动飞扬科技有限公司 | A kind of tire pressure detection method, device, storage medium and automobile |
| CN112488049A (en) * | 2020-12-16 | 2021-03-12 | 哈尔滨市科佳通用机电股份有限公司 | Fault identification method for foreign matter clamped between traction motor and shaft of motor train unit |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504743A (en) * | 2014-12-30 | 2015-04-08 | 深圳先进技术研究院 | Method and system for reconstructing internal region-of-interest image |
-
2016
- 2016-03-03 CN CN201610121292.3A patent/CN107153832B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504743A (en) * | 2014-12-30 | 2015-04-08 | 深圳先进技术研究院 | Method and system for reconstructing internal region-of-interest image |
Non-Patent Citations (2)
| Title |
|---|
| 杨亚明: "基于OpenCV的绝缘子及等电位线故障状态智能识别", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
| 杨红梅: "基于图像处理的接触网支持及悬挂装置不良状态检测", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108189626A (en) * | 2017-12-28 | 2018-06-22 | 深圳市灵动飞扬科技有限公司 | A kind of tire pressure detection method, device, storage medium and automobile |
| CN108189626B (en) * | 2017-12-28 | 2020-08-04 | 江苏灵动飞扬科技有限公司 | Tire pressure detection method and device, storage medium and automobile |
| CN112488049A (en) * | 2020-12-16 | 2021-03-12 | 哈尔滨市科佳通用机电股份有限公司 | Fault identification method for foreign matter clamped between traction motor and shaft of motor train unit |
| CN112488049B (en) * | 2020-12-16 | 2021-08-24 | 哈尔滨市科佳通用机电股份有限公司 | Fault identification method for foreign matter clamped between traction motor and shaft of motor train unit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107153832B (en) | 2019-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111080622B (en) | Neural network training method, workpiece surface defect classification and detection method and device | |
| Jeong et al. | Literature review and technical survey on bridge inspection using unmanned aerial vehicles | |
| CN109858367B (en) | Visual automatic detection method and system for worker through supporting unsafe behaviors | |
| US20140348415A1 (en) | System and method for identifying defects in welds by processing x-ray images | |
| Tyystjärvi et al. | Automated defect detection in digital radiography of aerospace welds using deep learning | |
| JP2019050376A5 (en) | ||
| CN115156093A (en) | Battery shell defect detection method, system and device | |
| CN107154033A (en) | A kind of high ferro contact net rotation ears vertical openings pin missing detection method and system | |
| Van Nguyen et al. | A mobile robot for automated civil infrastructure inspection and evaluation | |
| CN115035092A (en) | Image-based bottle detection method, device, equipment and storage medium | |
| CN107153832A (en) | A kind of high ferro contact net equipotential line releases detection method and system | |
| CN114693617A (en) | Method, system, equipment and storage medium for detecting defects of railway steel rails | |
| Park et al. | Patch-based crack detection in black box road images using deep learning | |
| CN107977953A (en) | Workpiece conductive features inspection method and workpiece conductive features check system | |
| CN103279823A (en) | Vehicle axle overhaul working procedure detecting method and system | |
| CN115761257A (en) | Method and device for detecting angle of part based on computer vision technology | |
| US20240169510A1 (en) | Surface defect detection model training method, and surface defect detection method and system | |
| CN107767372B (en) | Chip pin online visual detection system and method for layered parallel computing | |
| CN113192070B (en) | Weld image processing method and system based on computer vision | |
| KR20240064037A (en) | How to identify and characterize surface defects of objects and cracks in brake discs by artificial intelligence through fatigue testing | |
| CN113705567B (en) | Ship crack detection method, system, equipment and computer readable storage medium | |
| JP2011141202A (en) | Component inspection device and program | |
| Li et al. | Container damage identification based on Fmask-RCNN | |
| JP2011196982A (en) | Defect detection method, defect detection device, and program | |
| Cao et al. | Automated visual inspection of solar cell images using adapted morphological and edge detection algorithms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |