CN109615654A - Drainage pipeline inside corrosion depth and area measurement method based on binocular vision - Google Patents
Drainage pipeline inside corrosion depth and area measurement method based on binocular vision Download PDFInfo
- Publication number
- CN109615654A CN109615654A CN201910020768.8A CN201910020768A CN109615654A CN 109615654 A CN109615654 A CN 109615654A CN 201910020768 A CN201910020768 A CN 201910020768A CN 109615654 A CN109615654 A CN 109615654A
- Authority
- CN
- China
- Prior art keywords
- area
- corrosion
- point
- matching
- bus
- 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
- 238000005260 corrosion Methods 0.000 title claims abstract description 101
- 230000007797 corrosion Effects 0.000 title claims abstract description 98
- 238000000691 measurement method Methods 0.000 title claims abstract description 14
- 238000012937 correction Methods 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 13
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims abstract description 4
- 230000009466 transformation Effects 0.000 claims abstract description 4
- 239000013598 vector Substances 0.000 claims description 24
- 238000005259 measurement Methods 0.000 claims description 23
- 230000003628 erosive effect Effects 0.000 claims description 10
- 239000000284 extract Substances 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000007689 inspection Methods 0.000 claims description 6
- 238000013519 translation Methods 0.000 claims description 4
- 238000002474 experimental method Methods 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 2
- 230000009977 dual effect Effects 0.000 claims 1
- 238000001914 filtration Methods 0.000 claims 1
- 238000004088 simulation Methods 0.000 claims 1
- 230000001629 suppression Effects 0.000 claims 1
- 238000012360 testing method Methods 0.000 claims 1
- 230000002452 interceptive effect Effects 0.000 abstract description 3
- 238000010276 construction Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 11
- 230000000694 effects Effects 0.000 description 10
- 201000010099 disease Diseases 0.000 description 7
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 7
- 238000010586 diagram Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000004907 flux Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000003483 aging Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/22—Measuring arrangements characterised by the use of optical techniques for measuring depth
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/28—Measuring arrangements characterised by the use of optical techniques for measuring areas
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
- G06T7/85—Stereo camera calibration
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The present invention provides a kind of drainage pipeline inside corrosion depth and area measurement method based on binocular vision, include the following steps: by demarcating binocular camera, using gridiron pattern scaling board, using the graph outline of real corrosion of the pipe wall as reference, construction corrosion area is simulated in tube wall surface;Corrosion of the pipe wall area image is acquired using biocular systems, utilizes the nominal data distortion correction and three-dimensional correction of Bouguet algorithm and acquisition;Tube wall background and corrosion target are separated, it is final to obtain corrosion of the pipe wall region contour;The three-dimensional coordinate of match point is calculated based on principle of parallax;Using parallel lines affine transformation invariance, tube wall bus is calculated, using pipeline bus as the depth and area of auxiliary line computation corrosion area;The present invention has that detection is more intuitive, and interactive capability is strong, and detection is more automatic, fast, accurately feature, is conducive to staff and more preferably measures drainage pipeline inside corrosion depth and area.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of drainage pipeline inner surface based on binocular vision is rotten
Lose depth and area measurement method.
Background technique
Drainage pipeline is important one of the infrastructure in city, as the rapid development of sociaty and economy, the draining in city
Pipe-line system is gradually improved, and has been achieved for the achievement to attract people's attention and brings huge social benefit, while we also infuse
It anticipates and arrives, many pipeline agings are serious, and there are surface corrosions, in spite of illness operation, and thus bring hidden danger is to people's living standard and people
The influence of people's living safety is huge, detection of the development to drainage pipeline inside corrosion, in time with scientific method guidance
Maintenance work, has been the task of top priority.The disasters such as collapse to effectively prevent road surface, measurement drainage pipeline inside corrosion is deep
Degree and area, and progress pipeline rehabilitation becomes to be even more important in time.Currently, mainly dividing in engineering the measurement method of disease
To utilize laser range finder, impulse eddy current, Magnetic Flux Leakage Inspecting, structure light and CCTV method.
Drainage pipeline Defect inspection method, energy are completed using laser range finder, impulse eddy current, Magnetic Flux Leakage Inspecting, structure light
Disease three-dimensional data is effectively obtained, overcomes the problems, such as the uncertainty and work safety of artificial detection, but that there are measuring instruments is expensive,
Complicated for operation, the problems such as interactive capability is weak, affected by noise big;Piping disease is completed based on CCTV method to detect to obtain reality
The problems such as border application, detection is more intuitive, and interactive capability is strong, but not high there are detection accuracy.
Above-mentioned tube wall detection method can obtain tube wall point cloud data, but lack to the complete of corrosion of the pipe wall depth and area information
Automatic analysis method.Therefore, the drainage pipeline inside corrosion depth that set forth herein a kind of based on technique of binocular stereoscopic vision and
The measurement method of area, this method can be improved detection accuracy and real-time, and provide corrosion analysis strategy, so that detection is more
Add automatic, quick, accurate.
Summary of the invention
The purpose of the present invention is to provide a kind of drainage pipeline inside corrosion depth and the measurement methods of area, utilize
Biocular systems acquire tube wall image pair;By three-dimensional correction, contours extract, Feature Points Matching, corrosion area world coordinates is obtained;
Measurement auxiliary line finally parallel with bus using parallel lines reflection invariance foundation;Pass through point to auxiliary line distance and unique step
Triangle between auxiliary line is to calculating separately corrosion depth and area.It is compared by the measurement data and reference data of corrosion of the pipe wall,
Show that the set measurement method has good robustness and measurement accuracy.
In order to achieve the above object, the invention adopts the following technical scheme:
A kind of drainage pipeline inside corrosion depth and area measurement method based on binocular vision, comprising the following steps:
The S110 equipment preparation stage: building in-orbit pipeline crawler the right and left for camera, completes camera with ZHANGShi principle
Calibration, the internal reference matrix (M of left and right camera is acquired using OpenCVl,Mr), distortion factor vector (βl,βr), translation vector T, rotation
Steering volume R, re-projection matrix Q;
The S120 three-dimensional correction stage: pipeline inner wall corrosion area image is acquired using the left and right camera built, is utilized
Bouguet algorithm and S110 nominal data obtained carry out distortion correction and three-dimensional correction to the left images of acquisition;
S130 characteristic matching stage: to the left images handled by step S120, artificial Selecting All Parameters are utilized
Unrestrained water filling algorithm separation tube wall background and corrosion target, obtain the profile in corrosion of the pipe wall region;Profile point describes son with SURF
As matching variable, corrosion area describes son using FAST as matching variable, with Euclidean distance as detective operators, using BRIEF
It is most short to be used as matching condition, the profile point and corrosion area of left images are matched;Finally according to the matching knot of profile point
The matching result of fruit and erosion profile establishes confidence interval using hypothetical inspection using parallax as statistic, realizes that error hiding picks
It removes;
S140 three-dimensional coordinate calculation stages: three of match point in outline point and corrosion area are calculated based on principle of parallax
Tie up coordinate;
S150 corrosion of the pipe wall depth and area measurement stage: utilizing parallel lines affine transformation invariance, calculates tube wall bus,
Using tube wall bus as the depth and area of auxiliary line computation corrosion area;
Corrosion of the pipe wall depth and area measurement stage are divided into following two step in S150:
S151 tested point auxiliary line estimation: by the affine-invariant features of parallel lines it is found that being put down in three dimensions with axis
Capable several buses, the picture in image remain line and parallel to each other;There are two intersection point, this antinode exists for pipeline bus and profile
The vector V that image coordinate system is constitutedI, several intersection points to the constituted vector set of world coordinate system be VW={ (Xi,Yi,Zi)}。
VWIn each vector it is parallel to each other, by the property of parallel lines and related coefficient it is found that VWRelated coefficient between each dimension is 1, in parallel
The property of line and related coefficient are as follows: | ρ XY |=1 necessary and sufficient condition is that there are constant a, b, so that P { Y=a+bX }=1, that is, sit
Mark has synteny, then related coefficient absolute value is 1.Using above-mentioned parallel lines reflection invariance, if establishing parallel with axis
For fundatrix line as measurement auxiliary line, corrosion depth is minimum range of the measurement point to each auxiliary line.
The estimation of S152 tested point corrosion depth and tube wall area estimation: any bus g of tube walliWith the profile two of corrosion area
End intersection point (i, i') is in the vector of three-dimensional theorem in Euclid spaceCorrosion area any point coordinate p (x, y, z), opposite tube wall
Corrosion depth hp=min { di(p,gi), i ∈ { 0,2 ..., m ' }, di(p,gi) it is point p to straight line giDistance, m ' be bus
With the number of the intersection point of corrosion area profile, corroded area S, step are step-length.By corrosion area quadrangle area equivalent two
A triangle area, the then area of corrosion area are as follows:
S=sum { Sii′(i+step)+S(i+step)(i+step)′i′},i∈{1,1+step,1+2step,....},step∈N
Wherein Sii′(i+step)Indicate upper triangular area, S(i+step)(i+step)′i′Indicate that lower triangular area, N indicate
Natural number.
Detailed description of the invention
Present invention will be further explained below with reference to the attached drawings and examples:
Fig. 1 is the detail flowchart of inner wall of the pipe local erosion depth spot and area measurement based on binocular vision;
Fig. 2 is binocular imaging schematic diagram;
Fig. 3 is binocular graphic projection feature;
Fig. 4 is matching parallax distribution map;
Fig. 5 is that error hiding rejects parallax distribution map;
Fig. 6 is pipeline A Corrosion Disease model;
Fig. 7 is that bus projects to the picture on corrosion area image;
Fig. 8 is sample data figure;
Fig. 9 is three-dimensional correction effect picture and contours extract effect picture;
Figure 10 is erosion profile matching effect figure;
Figure 11 is corrosion area matching effect figure;
Figure 12 is bus projection search schematic diagram;
Figure 13 is pipeline bus Projection Analysis figure.
Table 1 is the relatively corrosive depth of characteristic point in the embodiment of the present invention;
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to be more clearly understood that the above objects, features and advantages of the present invention
The present invention is further elaborated, and following embodiment does not constitute a limitation of the invention.
Fig. 1 describe drainage pipeline inside corrosion depth of one of the embodiment of the present invention based on binocular vision and
The detailed process of area measurement method, the specific steps are as follows:
The S110 equipment preparation stage: this experiment purpose be verifying based on binocular vision drainage pipeline inside corrosion depth and
The feasibility of area measurement;Experiment uses 1920 × 1080 binocular cameras, using 9 × 6 gridiron pattern scaling board.With reality pipe
Wall erosion figure is referred to as profile, simulates construction corrosion area in tube wall surface.Fig. 8 shows erosion profile figure respectively, and final
Disease model.It is 319.47cm using criterion calculation paper measurement corroded area2。
Two cameras are built into in-orbit pipeline crawler the right and left respectively, camera calibration is completed with ZHANGShi principle.Two-phase
Machine shoots 20 width scaling board images respectively, and the internal reference matrix (M of left and right camera is acquired using OpenCVl,Mr), distortion factor vector
(βl,βr), translation vector T, rotating vector R, re-projection matrix Q, pixel error 0.595033.
The S120 three-dimensional correction stage: acquiring corrosion of the pipe wall area image using biocular systems, three-dimensional using Bouguet algorithm
Correction, Fig. 9 (a) show the three-dimensional correction effect picture using scaling board vision inspections Bouguet algorithm.
S130 characteristic matching stage: binocular Feature Points Matching feature: space arbitrary line l as shown in Figure 3, in S1、S2As flat
Face projection has rotational invariance.Face S0With two as plane S1、S2In parallel, straight line l is in plane S0Be projected as l0, l0In S1、S2Picture
Plane projection is l1,l2, by similar triangles property it is found that f/h=l1/l0=l2/l0, demonstrate,prove to obtain l1=l2.Therefore binocular image
With without considering scale and Rotation.The characteristics of based on the matching of above-mentioned binocular image, matching algorithm is rejected for rotation and ruler
The calculating process of degree can reduce match complexity, improve matching accuracy and speed.Characteristic matching stage specifically includes corrosion region
The sparse matching of domain contours extract, corrosion area profile point dense matching, the non-profile point of corrosion area and error hiding are rejected.
S131 corrosion area contours extract: image segmentation can be roughly divided into two classes from principle, and one is according to similar
Property, another kind is according to discontinuity.Discontinuity refers to the place that the mutation of those pixels occurs, and mostly occurs in object and object
The intersection of body.Because the inner vein of Corrosion Disease is considerably complicated, extracts corrosion outermost layer profile and use based on similitude
The effect of segmentation is barely satisfactory, and the edge of disease is found in the segmentation used herein based on discontinuity.
The matching of S132 corrosion area profile point: the matched accuracy of wire-frame image vegetarian refreshments directly affects later period Corrosion Evaluation, because
This selects to be suitble to the matching algorithm in corrosion of the pipe wall profile and region most important.Erosion profile is tube wall background and corrosion area
Intersection, background and corrosion area pixel space diversity factor are big, contrast is high.Therefore uncomfortable using the SAD algorithm of the sum of difference
It is matched with profile point;BRIEF description son be by the binary descriptor that compared pixels establish size in reference neighborhood, by
In above-mentioned erosion profile surrounding pixel characteristic of spatial distribution, will be showed in many dimensions with 0 or 1, loss partial dimensional is believed
Breath.Compared with foregoing description, the matching effect of SURF description is suitable with BRIEF, and the runing time of SURF is BRIEF
1/2.Therefore, SURF description is used to carry out the matching of corrosion area profile point as matching variable in the present embodiment.
The non-profile point matching of S133 corrosion area: in order to compensate for the precision of contours extract and three-dimensional correction to match point (xi,
yi) matching precision influence, increase right side point (x ' to be matchedi,y′i) search range (x, y) | | x-x 'i| < width, | y-y 'i
| < 1 }, wherein width indicates matching search width.Using the detection of FAST characteristic point and BRIEF description inside corrosion area
Son carries out the non-profile point matching of corrosion area as matching variable.
S134 error hiding is rejected: error hiding is inevitable in the matching process, and error hiding leads to our corrosion depth letters
Erroneous estimation is ceased, this has a significant impact to corrosion area assessment.It is representative by calculating corrosion area inside corrosion area
Point can be realized the regional assessment, therefore it is big to filter out Euclidean distance using the Euclidean distance between matching pair as condition for we
In the matching pair of threshold value;The matching result of the profile point of corrosion area is set using parallax as statistic using hypothetical inspection foundation
Believe section [+3 σ of μ -3 σ, μ], realizes that error hiding is rejected, μ, σ are respectively the mean value and variance of parallax.
Figure 10 is the erosion profile point matching effect figure using SURF description, wherein matches the width width=of search
2, show that the profile point for having chosen 1/2 is matched;Figure 11 is to describe son using FAST operator detection characteristic point and BRIEF to be
The matching effect figure inside the corrosion area of variable, FAST detection threshold value 20 are matched, Euclidean distance is Optimum Matching less than 15, special
It is 13805 that sign point, which extracts quantity, optimal number 584;Runing time is 2.756962s.
Fig. 4 (a) is respectively that SAD, BRIEF, SURF description match parallax distribution map, Fig. 4 (b) to Fig. 8 sample profile point
Backsight difference Butut is rejected for SURF error hiding, profile point is using the profile point of the top in image as starting point inverse time needle sort.Together
On, the width for matching search remains as width=2.SAD runing time and parallax variance are respectively 44.3163s, 54324;
BRIEF runing time and parallax variance are respectively 1.34548s, 15337;SURF runing time and parallax variance are respectively
0.628534s,21736.SAD runing time is longer, error hiding is more, and SURF describes sub- matching effect and describes sub- phase with BRIEF
When SURF runing time is BRIEF half.Using improved ORB and SURF algorithm corrosion area Feature Points Matching operation
Time is respectively as follows: 2.756962s, 3.2835s.Fig. 5 show respectively corrosion area using improve ORB obtain source matching and accidentally
Parallax distribution map with rejecting.
S140 three-dimensional coordinate calculation stages: the three-dimensional of match point is calculated based on principle of parallax using binocular stereo vision measurement
Coordinate.Binocular stereo vision measurement is based on principle of parallax, and Fig. 2 show binocular stereo imaging schematic diagram, the projection of twin camera
Center line distance is T, and target point P=(X, Y, Z) is respectively p in the image coordinate that left and right camera projectsl=(xl,yl) and pr=
(xr,yr);Parallax d=xl-xr;F is focal length;Camera coordinate system origin is located at left camera optical center, by Similar Principle of Triangle
Derive that three-dimensional coordinate of the P point under camera coordinate system is formula (2-1).
Using the above method, Figure 12 (b) pixel coordinate system is established, using ∠ AOB as the search model of pipeline bus in image
It encloses, 3 pixel coordinates of A, O, B are respectively (660,0), (960,1080), (1260,0), sampling step length 5 between A, B.
S150 utilizes parallel lines affine transformation invariance, calculates tube wall bus, rotten using pipeline bus as auxiliary line computation
Lose the depth and area in region;The depth of corrosion area and area measurement include following sub-step:
S161 tested point auxiliary line estimation: by the affine-invariant features of parallel lines it is found that parallel with axis in Fig. 6 (a)
Several buses remain line and parallel to each other in the picture as plane.If i, the world coordinates of i' is pi、pi', then pipeline bus with
The limited vector setes that profile intersection point is constitutedThereforeFor three-dimensional vector, V be collinearly to
Quantity set.By the property of related coefficient it is found that formula of correlation coefficient (4-2) absolute value between x, y, z | ρxy|=| ρxz|=| ρyz|
=1.
Fig. 7 is the photographs of Fig. 6 dummy model.Fig. 6 (a) median generatrix giPicture in Fig. 7 is li', n is bus in Fig. 7
The quantity of middle projecting direction.Due to pipeline bus giDirection is unknown, in order to determine pipeline bus picture direction, in image coordinate
Possible projecting direction { the l of n item is chosen in system within the scope of 0-360 degree1,l2,...,ln}.If lk∈{l1,l2,...,lnIt is it
In a possible projecting direction, then with lkParallel all straight line lk0...lkmWorld coordinates corresponding with profile both ends intersection point
Composed vector setCorresponding related coefficient isAs i=index, meetThen li'//lindexIndicate li' and lindexIn parallel, and ρindexFor ... ρindex-1,ρindex,
ρindex+1... local maximum.At this point, lindexFor the direction of the picture of the pipeline bus acquired.
Figure 12 (a) shows OA, the search sample graph of tri- direction buses of OD (960,0), OB projection respectively.Figure 13 (a) points
The corresponding ρ of 121 directions of search is not illustratedkScatter plot, Figure 13 (b) are shown with ρiFor determining bus perspective view, corresponding diagram 12
(b) OE (965,0) direction.Runing time 0.838884s.
The estimation of S152 tested point corrosion depth and tube wall area estimation: any bus g of tube walliWith the profile two of corrosion area
End intersection point (i, i') is in the vector of three-dimensional theorem in Euclid spaceCorrosion area any point coordinate p (x, y, z), opposite tube wall
Corrosion depth hp=min { di(p,gi), i ∈ { 0,2 ..., m ' }, di(p,gi) it is point p to straight line giDistance, m ' be bus
With the number of the intersection point of corrosion area profile, corroded area S, step are step-length.By corrosion area quadrangle area equivalent two
A triangle area, the then area of corrosion area are as follows:
S=sum { Sii′(i+step)+S(i+step)(i+step)′i′},i∈{1,1+step,1+2step,....},step∈N
Wherein Sii′(i+step)Indicate upper triangular area, S(i+step)(i+step)′i′Indicate that lower triangular area, N indicate
Natural number.
24 measurement points of uniform design in Fig. 9, measurement mean error are 5.5603mm, randomly select 8 measurement result column
Round and square central point is 24 measurement point positions in table 1, Figure 13 (c), rectangular for measurement point in table;Step=4 is rotten
Erosion area measurement is 358.157cm2。
Table 1
(mm)
Claims (4)
1. a kind of drainage pipeline inside corrosion depth and area measurement method based on binocular vision, which is characterized in that including
Following steps:
Step S110: building in-orbit pipeline crawler the right and left for camera, completes camera calibration with ZHANGShi principle, utilizes
OpenCV acquires the internal reference matrix (M of left and right cameral,Mr), distortion factor vector (βl,βr), translation vector T, rotating vector R, again
Projection matrix Q;
Step S120: pipeline inner wall corrosion area image is acquired using the left and right camera built, utilizes Bouguet algorithm and S110
Acquired nominal data carries out distortion correction and three-dimensional correction to the left images of acquisition;
Step S130: to the left images handled by step S120, the unrestrained water filling algorithm of artificial Selecting All Parameters is utilized
Tube wall background and corrosion target are separated, corrosion of the pipe wall region contour is obtained;Profile point using SURF describe son as match variable into
It goes and matches, corrosion area detects characteristic point using FAST as detective operators, using BRIEF description as matching variable progress
Match;Confidence area is finally established using hypothetical inspection using parallax as statistic according to the matching result of corrosion area profile point
Between, realize that error hiding is rejected;The matching result of the non-profile point of corrosion area, it is most short as matching condition using Euclidean distance, it filters out
Euclidean distance is greater than the matching pair of threshold value, realizes and rejects without matching;
Step S140: the three-dimensional coordinate of match point in outline point and corrosion area is calculated based on principle of parallax;
Step S150: utilizing parallel lines affine transformation invariance, calculates tube wall bus, rotten using pipeline bus as auxiliary line computation
Lose the depth and area in region;The depth of corrosion area and area measurement stage are divided into following two step:
S151 tested point auxiliary line estimation: by the affine-invariant features of parallel lines it is found that several buses parallel with axis, in picture
The picture of plane remains line and parallel to each other.The limited vector setes that pipeline bus and profile intersection point are constitutedIf
I, the world coordinates of i' is pi、pi',ThereforeFor three-dimensional vector, V is collinear vectors collection.
By the property of related coefficient it is found that formula of correlation coefficient (4-2) absolute value between x, y, z | ρxy|=| ρxz|=| ρyz|=1.
Bus giPicture be li', n is the quantity of bus projecting direction.Due to pipeline bus giDirection is unknown, in order to determine pipeline
Possible projecting direction { the l of n item is chosen in the direction of the picture of bus in image coordinate system within the scope of 0-360 degree1,l2,...,ln}。
If lk∈{l1,l2,...,lnBe a wherein possible projecting direction, then with lkParallel all straight line lk0...lkmWith profile
Vector set composed by the corresponding world coordinates of both ends intersection pointCorresponding related coefficient is
As i=index, meetThen li'//lindexIndicate li' and lindexIn parallel, and ρindexFor ...
ρindex-1,ρindex,ρindex+1... local maximum.At this point, lindexFor the direction of the picture of the pipeline bus found out.
The estimation of S152 tested point corrosion depth and tube wall area estimation: any bus g of tube walliIt is handed over the profile both ends of corrosion area
Point (i, i') is in the vector of three-dimensional theorem in Euclid spaceCorrosion area any point coordinate p (x, y, z), the corrosion of opposite tube wall
Depth hp=min { di(p,gi), i ∈ { 0,2 ..., m ' }, di(p,gi) it is point p to straight line giDistance, m ' is bus and rotten
The number of the intersection point of region contour is lost, corroded area S, step are step-length.Corrosion area quadrangle area is two three equivalent
Angular area, the then area of corrosion area are as follows:
S=sum { Sii′(i+step)+S(i+step)(i+step)′i′},i∈{1,1+step,1+2step,....},step∈N
Wherein Sii′(i+step)Indicate upper triangular area, S(i+step)(i+step)′i′Indicate that lower triangular area, N indicate nature
Number.
2. pipe surface corrosion depth according to claim 1 and area measurement method, which is characterized in that the step
Pipeline inner wall corrosion area image pair is acquired using biocular systems in S110, using 1920 × 1080 binocular cameras, using 9 ×
6 gridiron pattern scaling board constructs the corrosion of equal proportion in the surface simulation of experiment tube wall with reference to the etch pattern of real tube wall
Camera calibration is completed with ZHANGShi principle in region, and left and right camera is shot 20 width scaling board images respectively, acquired using OpenCV
Internal reference matrix (the M of left and right cameral,Mr), distortion factor vector (βl,βr), translation vector T, rotating vector R, re-projection matrix Q.
3. pipe surface corrosion depth according to claim 1 and area measurement method, which is characterized in that the step
S130 is specifically included:
Step S131: erosion profile is extracted;Inner wall of the pipe corrosion area picture noise is removed, gaussian filtering is used;Calculate gradient
Amplitude and direction;Non-maxima suppression is carried out to gradient magnitude;Edge is detected and connected with dual threashold value-based algorithm;
Step S132: SURF description is used to carry out the matching of the profile point of corrosion area as matching variable;
Step S133: in order to compensate for contours extract and three-dimensional correction to match point (xi,yi) matching precision influence, increase right side
Point (x ' to be matchedi,y′i) search range (x, y) | | x-x 'i| < width, | y-y 'i| < 1 }, wherein width indicates that matching is searched
Suo Kuandu;Matching is completed as matching variable using the detection of FAST characteristic point and BRIEF description inside corrosion area;
Step S134: Europe is filtered out using the Euclidean distance between matching pair as condition for the matching double points inside corrosion area
Family name's distance is greater than the matching pair of threshold value;The matching result of the profile point of corrosion area utilizes hypothetical inspection using parallax as statistic
It tests and establishes confidence interval [+3 σ of μ -3 σ, μ], wherein μ, σ are respectively the mean value and variance of parallax, realize that error hiding is rejected.
4. pipe surface corrosion depth according to claim 1 and area measurement method, which is characterized in that the step
S140 is specifically included:
The three-dimensional coordinate of match point is calculated based on principle of parallax using binocular stereo vision measurement, binocular stereo vision measurement is based on
Principle of parallax, the projection centre linear distance of twin camera are T, the image coordinate that target point P=(X, Y, Z) is projected in left and right camera
Respectively pl=(xl,yl) and pr=(xr,yr);Parallax d=xl-xr;F is focal length;Camera coordinate system origin is located at left video camera
Optical center derives three-dimensional coordinate of the P point under camera coordinate system by Similar Principle of Triangle are as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910020768.8A CN109615654B (en) | 2019-01-09 | 2019-01-09 | Method for measuring corrosion depth and area of inner surface of drainage pipeline based on binocular vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910020768.8A CN109615654B (en) | 2019-01-09 | 2019-01-09 | Method for measuring corrosion depth and area of inner surface of drainage pipeline based on binocular vision |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109615654A true CN109615654A (en) | 2019-04-12 |
CN109615654B CN109615654B (en) | 2020-01-21 |
Family
ID=66015668
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910020768.8A Active CN109615654B (en) | 2019-01-09 | 2019-01-09 | Method for measuring corrosion depth and area of inner surface of drainage pipeline based on binocular vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109615654B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110766669A (en) * | 2019-10-18 | 2020-02-07 | 南京大学 | Pipeline measuring method based on multi-view vision |
CN113393413A (en) * | 2020-03-11 | 2021-09-14 | 河海大学 | Water area measuring method and system based on monocular and binocular vision cooperation |
CN114562957A (en) * | 2022-02-28 | 2022-05-31 | 南京理工大学 | Stereoscopic vision industrial endoscope system based on double prisms and single camera and three-dimensional measurement method |
WO2022110043A1 (en) * | 2020-11-27 | 2022-06-02 | 西门子股份公司 | Erosion detection method and apparatus, and computer-readable medium |
CN115345879A (en) * | 2022-10-18 | 2022-11-15 | 济宁康盛彩虹生物科技有限公司 | Method for analyzing corrosion degree of inner wall of autoclave and predicting service life of autoclave based on image |
CN116758026A (en) * | 2023-06-13 | 2023-09-15 | 河海大学 | Dam seepage area measurement method based on binocular remote sensing image significance analysis |
CN117646828A (en) * | 2024-01-29 | 2024-03-05 | 中国市政工程西南设计研究总院有限公司 | Device and method for detecting relative displacement and water leakage of pipe jacking interface |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP4060334A1 (en) * | 2021-03-17 | 2022-09-21 | Tenaris Connections B.V. | A method and system for testing a pipe for an airbag system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102519969A (en) * | 2011-11-30 | 2012-06-27 | 北京市丰台区特种设备检测所 | Method for completely checking pressure pipeline with heat preservation function without stopping running |
CN107076676A (en) * | 2014-07-18 | 2017-08-18 | 迈确克斯过程控制公司 | Crack Detection and measurement in metallurgical tank |
CN206440400U (en) * | 2017-02-14 | 2017-08-25 | 天津健智者行空调技术有限公司 | A kind of central air-conditioning temperature sensor being applied under extremely warm environment |
US20180063370A1 (en) * | 2016-08-30 | 2018-03-01 | Konica Minolta, Inc. | Image processing apparatus, image forming apparatus, image forming system and image processing method |
RU2646525C1 (en) * | 2016-11-21 | 2018-03-05 | Общество с ограниченной ответственностью "Научно-исследовательский институт природных газов и газовых технологий - Газпром ВНИИГАЗ" | Method of determining the parameters of fragmentation wound in accidents at facilities handling compressed gas |
CN108680492A (en) * | 2016-04-29 | 2018-10-19 | 天津大学 | The assay method of corrosion depth in the galvanic corrosion of metal welding joints position |
CN109087270A (en) * | 2018-09-04 | 2018-12-25 | 中国矿业大学(北京) | One kind being based on improved convolution match tracing pipe video image defogging Enhancement Method |
-
2019
- 2019-01-09 CN CN201910020768.8A patent/CN109615654B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102519969A (en) * | 2011-11-30 | 2012-06-27 | 北京市丰台区特种设备检测所 | Method for completely checking pressure pipeline with heat preservation function without stopping running |
CN107076676A (en) * | 2014-07-18 | 2017-08-18 | 迈确克斯过程控制公司 | Crack Detection and measurement in metallurgical tank |
CN108680492A (en) * | 2016-04-29 | 2018-10-19 | 天津大学 | The assay method of corrosion depth in the galvanic corrosion of metal welding joints position |
US20180063370A1 (en) * | 2016-08-30 | 2018-03-01 | Konica Minolta, Inc. | Image processing apparatus, image forming apparatus, image forming system and image processing method |
RU2646525C1 (en) * | 2016-11-21 | 2018-03-05 | Общество с ограниченной ответственностью "Научно-исследовательский институт природных газов и газовых технологий - Газпром ВНИИГАЗ" | Method of determining the parameters of fragmentation wound in accidents at facilities handling compressed gas |
CN206440400U (en) * | 2017-02-14 | 2017-08-25 | 天津健智者行空调技术有限公司 | A kind of central air-conditioning temperature sensor being applied under extremely warm environment |
CN109087270A (en) * | 2018-09-04 | 2018-12-25 | 中国矿业大学(北京) | One kind being based on improved convolution match tracing pipe video image defogging Enhancement Method |
Non-Patent Citations (2)
Title |
---|
JIANG CHUNLEI.ET.AL.: "The research of natural gas pipeline leak detection based on adaptive filter technology", 《 PROCEEDINGS OF 2013 2ND INTERNATIONAL CONFERENCE ON MEASUREMENT, INFORMATION AND CONTROL》 * |
李涛涛等: "基于多视觉线结构光传感器的大尺度测量方法", 《中国激光》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110766669B (en) * | 2019-10-18 | 2022-06-21 | 南京大学 | Pipeline measuring method based on multi-view vision |
CN110766669A (en) * | 2019-10-18 | 2020-02-07 | 南京大学 | Pipeline measuring method based on multi-view vision |
CN113393413A (en) * | 2020-03-11 | 2021-09-14 | 河海大学 | Water area measuring method and system based on monocular and binocular vision cooperation |
CN113393413B (en) * | 2020-03-11 | 2022-09-13 | 河海大学 | Water area measuring method and system based on monocular and binocular vision cooperation |
WO2022110043A1 (en) * | 2020-11-27 | 2022-06-02 | 西门子股份公司 | Erosion detection method and apparatus, and computer-readable medium |
CN114562957B (en) * | 2022-02-28 | 2024-05-17 | 南京理工大学 | Industry endoscope system based on biprism single-camera stereoscopic vision and three-dimensional measurement method |
CN114562957A (en) * | 2022-02-28 | 2022-05-31 | 南京理工大学 | Stereoscopic vision industrial endoscope system based on double prisms and single camera and three-dimensional measurement method |
CN115345879A (en) * | 2022-10-18 | 2022-11-15 | 济宁康盛彩虹生物科技有限公司 | Method for analyzing corrosion degree of inner wall of autoclave and predicting service life of autoclave based on image |
CN115345879B (en) * | 2022-10-18 | 2023-03-24 | 济宁康盛彩虹生物科技有限公司 | Method for analyzing corrosion degree of inner wall of autoclave and predicting service life of autoclave based on image |
CN116758026A (en) * | 2023-06-13 | 2023-09-15 | 河海大学 | Dam seepage area measurement method based on binocular remote sensing image significance analysis |
CN116758026B (en) * | 2023-06-13 | 2024-03-08 | 河海大学 | Dam seepage area measurement method based on binocular remote sensing image significance analysis |
CN117646828A (en) * | 2024-01-29 | 2024-03-05 | 中国市政工程西南设计研究总院有限公司 | Device and method for detecting relative displacement and water leakage of pipe jacking interface |
CN117646828B (en) * | 2024-01-29 | 2024-04-05 | 中国市政工程西南设计研究总院有限公司 | Device and method for detecting relative displacement and water leakage of pipe jacking interface |
Also Published As
Publication number | Publication date |
---|---|
CN109615654B (en) | 2020-01-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109615654A (en) | Drainage pipeline inside corrosion depth and area measurement method based on binocular vision | |
Liu et al. | The use of laser range finder on a robotic platform for pipe inspection | |
CN108932475A (en) | A kind of Three-dimensional target recognition system and method based on laser radar and monocular vision | |
CN108335350A (en) | The three-dimensional rebuilding method of binocular stereo vision | |
CN109615611A (en) | A kind of insulator self-destruction defect inspection method based on inspection image | |
CN107680137B (en) | A kind of polyphaser calibration and parameter optimization method based on sphere projection common tangent | |
CN103988226A (en) | Method for estimating camera motion and for determining three-dimensional model of real environment | |
CN106504189A (en) | A kind of ocean disaster area extracting method based on GIS | |
CN114067197B (en) | Pipeline defect identification and positioning method based on target detection and binocular vision | |
CN109711321A (en) | A kind of wide Baseline Images unchanged view angle linear feature matching process of structure adaptive | |
CN113393439A (en) | Forging defect detection method based on deep learning | |
CN106125907A (en) | A kind of objective registration method based on wire-frame model | |
Kim et al. | The effective acquisition and processing of 3D photogrammetric data from digital photogrammetry for construction progress measurement | |
CN106959101B (en) | A kind of indoor orientation method based on optical flow method | |
CN109598755A (en) | Harmful influence leakage detection method based on binocular vision | |
Ziqiang et al. | Research of the algorithm calculating the length of bridge crack based on stereo vision | |
CN107808160B (en) | Three-dimensional building extraction method and device | |
Haala et al. | Combining Laser Scanning and Photogrammetry-A Hybrid Approach for Heritage Documentation. | |
CN104765702B (en) | A kind of type face data acquisition method in ship plank extreme misery process | |
CN109308706B (en) | Method for obtaining three-dimensional curved surface area through image processing | |
Zhang et al. | Building extraction from high-resolution remote sensing images based on GrabCut with automatic selection of foreground and background samples | |
CN106097362A (en) | Automatically detection and the localization method of artificial circular labelling in a kind of x-ray image | |
Zou et al. | Flood Depth Assessment with Location-Based Social Network Data and Google Street View-A Case Study with Buildings as Reference Objects | |
Hart et al. | Automated pipeline reconstruction using deep learning & instance segmentation | |
CN111046906B (en) | Reliable encryption matching method and system for planar feature points |
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 |