CN110009610A - Visual detection method for surface damage of reservoir dam protection slope and bionic device - Google Patents
Visual detection method for surface damage of reservoir dam protection slope and bionic device Download PDFInfo
- Publication number
- CN110009610A CN110009610A CN201910235217.3A CN201910235217A CN110009610A CN 110009610 A CN110009610 A CN 110009610A CN 201910235217 A CN201910235217 A CN 201910235217A CN 110009610 A CN110009610 A CN 110009610A
- Authority
- CN
- China
- Prior art keywords
- bionical
- crackle
- pit
- image
- damage
- 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.)
- Pending
Links
- 230000006378 damage Effects 0.000 title claims abstract description 68
- 238000001514 detection method Methods 0.000 title claims abstract description 44
- 230000004224 protection Effects 0.000 title claims abstract description 43
- 239000011664 nicotinic acid Substances 0.000 title claims abstract description 12
- 230000000007 visual effect Effects 0.000 title abstract description 6
- 208000037656 Respiratory Sounds Diseases 0.000 claims abstract description 45
- 239000004568 cement Substances 0.000 claims abstract description 21
- 241000196324 Embryophyta Species 0.000 claims description 24
- 238000000034 method Methods 0.000 claims description 15
- 238000003708 edge detection Methods 0.000 claims description 14
- 238000012937 correction Methods 0.000 claims description 12
- 230000009466 transformation Effects 0.000 claims description 12
- 208000027418 Wounds and injury Diseases 0.000 claims description 10
- 208000014674 injury Diseases 0.000 claims description 10
- 230000006735 deficit Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 239000000463 material Substances 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 235000013162 Cocos nucifera Nutrition 0.000 claims description 3
- 244000060011 Cocos nucifera Species 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 2
- 230000014759 maintenance of location Effects 0.000 claims description 2
- 230000007246 mechanism Effects 0.000 claims description 2
- 230000000717 retained effect Effects 0.000 claims description 2
- 230000001629 suppression Effects 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 2
- 238000005259 measurement Methods 0.000 abstract 1
- 230000001681 protective effect Effects 0.000 abstract 1
- 230000008859 change Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- NJPPVKZQTLUDBO-UHFFFAOYSA-N novaluron Chemical compound C1=C(Cl)C(OC(F)(F)C(OC(F)(F)F)F)=CC=C1NC(=O)NC(=O)C1=C(F)C=CC=C1F NJPPVKZQTLUDBO-UHFFFAOYSA-N 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 206010037660 Pyrexia Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001314 paroxysmal effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- 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
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B25/00—Models for purposes not provided for in G09B23/00, e.g. full-sized devices for demonstration purposes
- G09B25/04—Models for purposes not provided for in G09B23/00, e.g. full-sized devices for demonstration purposes of buildings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8874—Taking dimensions of defect into account
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Immunology (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Pathology (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a visual detection method and a bionic device for damage to the surface of a protective slope of a reservoir dam. Reservoir dam bank protection surface damage visual detection's bionical device includes bionical cement tree and two miniature cameras promptly of two mesh stereo vision systems, and the middle part of bionical cement tree is provided with bionical aigrette head, and the overhead at bionical aigrette is installed to the mesh stereo vision system, and bionical aigrette head can be through rotating the angle of adjusting the mesh stereo vision system, bank protection damage such as crackle or pit on real-time supervision reservoir dam bank protection surface. The invention has the advantages of low cost, simple structure and high detection precision, can detect the surface damage condition of the dam protection slope in real time, and saves technical measurement personnel.
Description
Technical field
The present invention relates to detection technique field, in particular to a kind of reservoir dam slope protection surface damage visible detection method and
Bionic device.
Background technique
Reservoir dam multidigit is ringed on three sides by mountains in mountain area, and the crackle or collapse equivalent damage of dam slope protection are Major Diseases
One of, once the natural calamities such as Heavy Rain of Typhoon or unpredictable earthquake occur, damage, which causes reservoir to breach a dyke, to bring to the people
It seriously endangers, especially paroxysmal collapse is most dangerous.It is early therefore, it is necessary to periodically be detected to slope protection health status
Phase discovery and lesion assessment are of great significance with disaster alarm.Traditional artificial periodic detection crackle and its damage method operation
When it is dangerous, detection accuracy is very low, because reservoir is located at mountain area, have inconvenient traffic, dam health detection and not in time, generation
Crack or small pit are found not in time and reparation causes seam the inside also to grow weeds, make crackle or injury region Quick Extended.Though
So the damage check based on image and cement fissure Study of recognition have formed hot spot, but reservoir dam slop protection material has mixed mud
With the different materials such as stone masonry, in addition weeds background, regularly slope protection crackle is difficult aircraft vision in telemeasurement gradient
Guarantee precision.Therefore, there is an urgent need to study the detection of the surface damages such as reservoir dam slope protection crackle and pit under complex background
Method.
Summary of the invention
It is an object of the invention to overcome disadvantage existing in the prior art, provide a kind of detection accuracy it is high, it is at low cost, knot
The simple reservoir dam slope protection surface damage visible detection method of structure and bionic device.
The purpose of the invention is achieved by the following technical solution:
A kind of reservoir dam slope protection surface damage visible detection method, includes the following steps:
(1) it camera calibration: firstly, being demarcated respectively to two video cameras, obtains and is used for correcting captured pattern distortion
Intrinsic parameters of the camera and distortion parameter；The calibration for carrying out stereoscopic vision, obtain positional relationship between two video cameras and
For carrying out the re-projection matrix of three-dimensionalreconstruction；It has been secured since two video cameras are installed rear relative position, it is above-mentioned
Parameter will not all change again, therefore the step only need to start to carry out once；
(2) it obtains correction image: capturing the digital picture of comprehensive crackle or pit by video camera, then basis is taken the photograph
The calibration result of camera carries out distortion correction to digital picture, obtains the correction image of crackle or pit；
(3) gray processing and smooth: the correction image of crackle or pit is subjected to gray processing, by tri- channels its R, G, B, often
A channel is weighted and averaged with different weights；Then detection zone is smoothed with Gaussian function, obtains crackle
Or the smooth single channel gray level image of pit；
(4) it improves Canny edge detection: the marginal information of the smooth single channel gray level image of crackle or pit is examined
It surveys, using improved Canny edge detection algorithm, the selection of high threshold H and Low threshold L is carried out according to the global feature of image,
So that the marginal information detected is relatively more and continuous, to detect the complete edge information of crackle or pit；
(5) it is based on Noise Elimination from Wavelet Transform: noise like of cutting weeds further is gone based on wavelet transformation Canny algorithm, for containing
There is the complex background of weeds noise like, first establish a kernel function φ (x, y), then carry out edge detection, if kernel function φ
The integral of (x, y) is 0, then to x, the derivative in y both direction does wavelet；Then to the crackle for eliminating weeds noise like
Or the surface of pit carries out three-dimensionalreconstruction；
(6) Stereo matching: obtain correction image to and the noise like that goes to cut weeds after, the left and right of left and right cameras is schemed
As carrying out Stereo matching, the disparity map of each match point is obtained；
(7) three-dimensionalreconstruction: after obtaining the disparity map of step (6), then the re-projection matrix obtained according to step (1), it generates
Three-dimensional point cloud carries out three-dimensionalreconstruction；
(8) assess damage type and degree: according to the different shape of damage, the method that can select respectively geometry teaching is intended
Close crackle or length, width or the approximate diameter of pit, depth parameter；Then it is assessed, obtains degree of injury and damage shape
Related data is finally fed back to user by formula.
In step (1), the video camera is wide-angle 16 channel thermal cameras of high-precision.
In step (3), the weight of gray processing weighting is using classical value disclosed in document；Smoothly located using Gaussian function
It manages, the number in the core of Gaussian filter is that Gaussian Profile is presented, this is different from mean filter, the core of mean filter
Each value be it is equal, therefore, the information of crackle or pit in original image can be more preferably obtained using Gaussian filter.
In order to eliminate noise jamming and not influence the information of damage field, when carrying out picture smooth treatment, the standard of Gaussian function
The selection of variances sigma is most important.To overcome classical Canny algorithm using the defect that the parameter is manually set, in N × N window,
Using variance as one of the module for choosing σ, topography's information in window has thus been taken into account.And non-edge area in image
Domain is often the lesser pixel of variance by the lesser region of noise pollution, can be using these pixels as marginal point and noise spot
Reference.In view of the overall permanence of image, using minimum variance as the module for choosing σ value, wherein M, E, σ are N × N respectively
Mean value, variance and Gauss standard variance in window；EminIt is minimum variance, when variance minimum, σ minimum value is 1；
Emin=min (E) (3)
No matter variance E value is very big or very little, σ value will be less than the value of formula (4) calculating, can prevent figure so in this way
As excess smoothness or edge fog.
In step (4), when carrying out edge detection with Canny algorithm, the acquisition of high-low threshold value is to detecting that edge believes
Breath number and detection correctness with decisive action；Since the starting point of edge detection is often controlled by high threshold, if high
Threshold value is smaller, although marginal information retains more, error detection can also become very much；If it is very big to define high threshold, though
Right error detection can become seldom, but will cause target critical information loss.Therefore, it is calculated using improved Canny edge detection
Method selects to consider image overall permanence when high-low threshold value, the marginal information detected in this way will it is relatively more and be it is continuous, from
And detect the information in crack in whole image.
In step (4), the selection of high threshold H and Low threshold L are carried out using improved Canny edge detection algorithm, are calculated
Method is as follows；
Wherein Eave、fave、lw、lhIt is the height of average variance, average gray, the width of image, image respectively；It is wherein low
Threshold value L=H/2, M refer to the mean value in N × N window, k=1/M, EminIt is minimum variance, fm(i, j) is in the picture i-th
The pixel value of row, jth column position.M is represented from 1 in summation symbol and is added to lw*lh。
In step (5), wavelet transformation Canny algorithm is as follows: the complex background containing weeds for removing crack image,
A kernel function φ (x, y) is first established, then carries out edge detection, if the integral of kernel function φ (x, y) is 0, then to x, y two
Derivative on direction does wavelet, and S is the area that x and y are surrounded in formula:
In order to remove noise of cutting weeds, according to the principle of wavelet transformation, the filter of a wavelet transformation is made, if scale is
2j, discrete wavelet transformation is carried out to image, component is the Local modulus maxima of wavelet transformation modulus value on gradient direction, such as formula
(9) shown in:
Shown in modulus value such as formula (10):
Shown in gradient direction such as formula (11):
If (x, y) is its local maximum, when local maxima, declines along gradient direction, can be in the image detection of crack
Modulus value is obtained, for retaining its crack information maximum value, while going the non-maximum of information of cutting weeds；Greater than the local maximum of threshold value
Value could be retained as edge；Specific method is to there is the image of weeds to carry out non-maximum suppression using wavelet module value
It makes, the background pixel point other than fracture region is inhibited, then obtains the threshold value of weeds in the picture in proportion, if weeds
Maximum be less than given threshold value and finally come out endpoint detections with Canny then just will be deleted weeds.
In step (6), the principle of Stereo matching is to choose the matching basic unit such as point, line, surface, by it along corrected
Image afterwards traverses the horizontal direction of (image is to two pictures for referring to the same time acquisition of two video cameras in left and right)
(traversal refers to along certain search pattern, successively does to each pixel and once and only does primary access) search, according to matching
The similarity determination of basic unit is matching or mismatches, to seek the matching relationship of pixel in left images；Due to a left side
The visual field of right video camera is different, and match point is inevitable different in the position of two video cameras, therefore it is possible thereby to calculates corresponding view
Difference figure；Core concepts of the solid four with algorithm be make full use of first basic constraint condition eliminate more solutions in matching process,
Then ambiguity problem converts energy function optimization problem for matching problem；The solution form of Stereo Matching Algorithm is to pass through
The cost energy function based on Matching unit is established, different matching strategies is taken to make this cost energy function minimum
Change to estimate pixel parallax value.Into after crossing step (2) to step (5) processing, the non-targeted part of left images is gone
It removes, only remaining fragmented parts, therefore the parallax obtained in Stereo matching also only has fragmented parts.
In step (7), the generation method of three-dimensional point cloud is the principle of triangulation using similar triangles, the depth of certain point
There are following relationships between degree and its parallax:Wherein z is the depth distance of certain point, and d is parallax, and f is focal length, b
For the parallax range between two cameras；Disparity map in step (6) is actually the parallax collection of each pixel on image
It closes, therefore, the three-dimensional point cloud of target can be generated；Three-dimensional point cloud is the set of target surface three-dimensional coordinate point, from intuitive angle
It sees, if there is damages for dam, the part and dam surface be not in a depth.
In step (8), crackle is fitted using nonlinear curve；For pit, selected according to its image analysis oval
Or circle is fitted.Rule of thumb, crackle has a long and thin feature, and length is much larger than width, and the length and width gap of pit
It is not too large, therefore the length-width ratio of crackle or pit has been predefined, according to experience early period, presetting length-width ratio is 10, works as image
When the length-width ratio of upper damage is greater than the value, it is judged as crackle；It is on the contrary then be judged as pit.For crackle, using polynomial curve
Its three-dimensional point cloud is fitted, to solve length, and width information then pass through calculate both sides of the edge point in three-dimensional point cloud away from
From obtaining.For pit, the boundary of its three-dimensional point cloud is extracted first, is then carried out boundary point using least square method oval quasi-
It closes, obtains length of the elliptical long side as pit, width of the elliptical short side as pit, the depth of pit then passes through three-dimensional
Point cloud pits obtain at a distance from dam surface.
In step (8), a set of degree of injury judgment basis first is established to crackle and pit respectively: for crackle, with its length
Degree and width are foundation, divide impairment scale step by step according to length and width size, short and thin crackle grade is minimum, long and thick
Crackle grade highest；Impairment scale is divided step by step according to size and the depth using its size and the depth as foundation for pit,
Small and shallow pit grade impairment scale is minimum, and big and deep Notch damage grade is maximum；Obtain the size letter of crackle or pit
After breath, compared with the impairment scale judgment basis in backstage, so that it is determined that degree of injury.
A kind of bionic device of reservoir dam slope protection surface damage vision-based detection, including bionical cement tree 5 and binocular solid
Vision system i.e. two microcam 7, using above-mentioned reservoir dam slope protection surface damage visible detection method；Bionical cement tree
5 middle part is provided with bionical egression head 8, and Binocular Stereo Vision System is mounted on bionical egression head, and bionical egression head can lead to
Rotation is crossed to adjust the angle of Binocular Stereo Vision System, the shield such as the crackle or pit on 10 surface of real-time monitoring reservoir dam slope protection
Slope damage 9.
The top of bionical cement tree 5 is equipped with artificial tree top cover 2, and the periphery of artificial tree top cover 2 passes through retention mechanism
It is fixed with solar panels and its battery 1.On the one hand solar panels can receive solar energy source, pass through electric wire 3 and microcam 7
It is connected, provides the energy for microcam；Another aspect solar panels can keep off the rain for camera and the sun, and it is straight to reduce sunlight
According to avoid camera fever from damaging reduces with precision.
Microcam 7 is connect by cable and its transmitter 4 with computer 14, by the data transmission of microcam 7
To computer 14.The related Control System of reservoir dam slope protection surface damage visible detection method and soft is installed in computer 14
Part system, including vision control system and software systems, surface damage visual detection algorithm, speech prompting system.
Bionical cement tree 5 is to imitate coastal Anti-Typhoon cocoanut tree structure, and the cross section of bionical cement tree 5 is approximate ellipse
Shape is provided with annual ring groove, which can evacuate power and reduce resistance and vibration.
Bionical egression head 8 is installed on bionical cement tree 5 by screw, miniature to adjust by adjusting screw
The angle of video camera.It is provided with camera mounting bracket on bionical egression head 8, for installing microcam 7.Bionical egression head 8
Top is provided with bionical leaf 6, is lubricious material, for being 7 rain cover of microcam and sunlight.
Bionical cement tree 5 is fixed on artificial tree pedestal 11 by nut 12 and installation screw rod 13.
The bionic device of above-mentioned reservoir dam slope protection surface damage vision-based detection passes through miniature video camera at work first
Machine acquisition damage target image, and real-time transmission carries out image procossing by computer, analysis is to remote computer based system
No there are damage phenomenons, make corresponding prompt, degree of injury and health evaluating to technical staff.
The present invention has the following advantages that compared with prior art and effect:
(1) cost of bionic device of the present invention is low, and structure is simple, can be with the surface damage feelings of real-time detection dam slope protection
Condition, saving technique survey crew；
(2) precision of detection method is high, passes through binocular vision system, human-computer interaction, noise remove, three-dimensionalreconstruction
Deng can intuitively detect dam injured surface situation；
(3) present invention facilitates operation, degree of impairment can be sent to dam by way of voice, short message or wechat
Administrative staff.
Detailed description of the invention
Fig. 1 is schematic three dimensional views when apparatus of the present invention use.
Fig. 2 is the bionical egression head schematic diagram of apparatus of the present invention.
Fig. 3 is algorithm flow chart of the invention.
1, solar panels and its battery；2, artificial tree top cover；3, electric wire；4, cable and its transmitter；5, bionical cement
Tree；6, bionical leaf, 7, microcam；8, bionical egression head；9, slope protection damages；10, slope protection；11, artificial tree pedestal；12,
Nut；13, screw rod is installed；14, computer.
Specific embodiment
Further detailed description is done to the present invention below with reference to embodiment, embodiments of the present invention are not limited thereto.
Embodiment 1
As shown in Figure 1, a kind of reservoir dam slope protection surface damage vision-based detection bionic device, including bionical 5 He of cement tree
Binocular Stereo Vision System i.e. two microcam 7, using reservoir dam slope protection surface damage visible detection method；Bionical water
The middle part of mud tree 5 is provided with bionical egression head 8, and Binocular Stereo Vision System is mounted on bionical egression head, and bionical egression head can
To adjust the angle of Binocular Stereo Vision System, the crackle or pit on 10 surface of real-time monitoring reservoir dam slope protection by rotating
Equal slope protections damage 9.Bionical cement tree 5 is to imitate coastal Anti-Typhoon cocoanut tree structure, and the cross section of bionical cement tree 5 is approximate ellipse
Circle is provided with annual ring groove, which can evacuate power and reduce resistance and vibration.Bionical egression head 8 passes through screw
It is installed on bionical cement tree 5, the angle of microcam is adjusted by adjusting screw.It is arranged on bionical egression head 8
There is camera mounting bracket, for installing microcam 7.The top of bionical egression head 8 is provided with bionical leaf 6, is lubricious material,
For for 7 rain cover of microcam and sunlight.As shown in Fig. 2, two microcams 7 are mounted on bionical cement tree 5 one
The both sides of the head for determining the bionical egression on mounting rack intermediate in height has angle square correction, and 2 video cameras are needed against slope protection,
Guarantee that camera lens center line is vertical with slope protection, and as a pair of eyes, the image that video camera obtains is given by network transmission
The software systems of remote computer visual pattern processing.
As shown in figure 3, first being demarcated to video camera when carrying out reservoir dam slope protection surface damage vision-based detection, lead to
It crosses image preprocessing: including going to distort to the image for obtaining damage, correcting, obtain the correction image of injury region；Gray processing and flat
It is sliding；Then detection zone is smoothed with Gaussian function, obtains the smooth single channel gray level image of crackle；With improved
Canny edge detection algorithm extracts damage edge, is based on Noise Elimination from Wavelet Transform, further removes noise, such as weeds class is made an uproar
Sound.Three-dimensionalreconstruction is carried out to injured surface again: Stereo matching being carried out to the image of left and right camera, generates three-dimensional point cloud, further
Noise is removed, three-dimensionalreconstruction is carried out.Identification of damage is that crackle or pit can select several respectively according to the different shape of damage
The method of what mathematics is fitted crackle or length, width or the approximate diameter of pit, depth parameter.Final system is to these parameters
It is assessed, obtains degree of injury and damage type, provide decision data to repair slope protection.Vision software should include voice system
System prompts technology or operator.
The above description is only an embodiment of the present invention, but embodiment of the present invention are not limited by the above embodiments,
It is other it is any without departing from the spirit and principles of the present invention made by changes, modifications, substitutions, combinations, simplifications, be
The substitute mode of effect, is included within the scope of the present invention.
Claims (10)
1. a kind of reservoir dam slope protection surface damage visible detection method, it is characterised in that include the following steps:
(1) camera calibration: firstly, demarcating respectively to two video cameras, taking the photograph for correcting captured pattern distortion is obtained
Camera inner parameter and distortion parameter；The calibration for carrying out stereoscopic vision, obtains the positional relationship between two video cameras and is used for
Carry out the re-projection matrix of three-dimensionalreconstruction；
(2) it obtains correction image: the digital picture of comprehensive crackle or pit is captured by video camera, then according to video camera
Calibration result to digital picture carry out distortion correction, obtain the correction image of crackle or pit；
(3) gray processing and smooth: the correction image of crackle or pit is subjected to gray processing, by tri- channels its R, G, B, Mei Getong
Road is weighted and averaged with different weights；Then detection zone is smoothed with Gaussian function, obtains crackle or recessed
The smooth single channel gray level image in hole；
(4) it improves Canny edge detection: the marginal information of the smooth single channel gray level image of crackle or pit is detected,
Using improved Canny edge detection algorithm, the selection of high threshold H and Low threshold L are carried out according to the global feature of image, so that
The marginal information detected is relatively more and continuous, to detect the complete edge information of crackle or pit；
(5) it is based on Noise Elimination from Wavelet Transform: noise like of cutting weeds further is gone based on wavelet transformation Canny algorithm, for containing miscellaneous
The complex background of careless noise like first establishes a kernel function φ (x, y), then carries out edge detection, if kernel function φ (x, y)
Integral be 0, then to x, the derivative in y both direction does wavelet；Then to the crackle that eliminates weeds noise like or recessed
The surface in hole carries out three-dimensionalreconstruction；
(6) Stereo matching: obtain correction image to and the noise like that goes to cut weeds after, to the left images of left and right cameras into
Row Stereo matching obtains the disparity map of each match point；
(7) three-dimensionalreconstruction: after obtaining the disparity map of step (6), then the re-projection matrix obtained according to step (1), it generates three-dimensional
Point cloud, carries out three-dimensionalreconstruction；
(8) assess damage type and degree: according to the different shape of damage, the method for selecting geometry teaching respectively is fitted crackle
Or length, width or the approximate diameter of pit, depth parameter；Then it is assessed, obtains degree of injury and damage type, finally
Related data is fed back into user.
2. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (3)
In, it is smoothed using Gaussian function, the number in the core of Gaussian filter is that Gaussian Profile is presented, in order to eliminate
Noise jamming and the information for not influencing damage field, when carrying out picture smooth treatment, the choosing of the standard variance σ of Gaussian function
Taking is overall permanence in view of image, and using minimum variance as the module for choosing σ value, wherein M, E, σ are N × N window respectively
Mean value, variance and Gauss standard variance in mouthful；EminIt is minimum variance, when variance minimum, σ minimum value is 1；
Emin=min (E) (3)
3. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (4)
In, the selection of high threshold H and Low threshold L are carried out using improved Canny edge detection algorithm, algorithm is as follows；
Wherein Eave、fave、lw、lhIt is the height of average variance, average gray, the width of image, image respectively；Wherein Low threshold L
=H/2, M refer to the mean value in N × N window, k=1/M, EminIt is minimum variance, fm(i, j) is the i-th row, jth in the picture
The pixel value of column position.
4. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (5)
In, wavelet transformation Canny algorithm are as follows: for removing the complex background containing weeds of crack image, first establish a kernel function
φ (x, y), then carry out edge detection, if the integral of kernel function φ (x, y) is 0, then to x, the derivative in y both direction does base
This small echo, S is the area that x and y are surrounded in formula:
In order to remove noise of cutting weeds, according to the principle of wavelet transformation, the filter of a wavelet transformation is made, if scale is 2j, right
Image carries out discrete wavelet transformation, and component is the Local modulus maxima of wavelet transformation modulus value on gradient direction, such as formula (9) institute
Show:
Shown in modulus value such as formula (10):
Shown in gradient direction such as formula (11):
If (x, y) is its local maximum, when local maxima, declines along gradient direction, can obtain in the image detection of crack
Modulus value for retaining its crack information maximum value, while going the non-maximum of information of cutting weeds；Greater than the local maximum ability of threshold value
It can be retained as edge；Specific method be to have the image of weeds using wavelet module value carry out non-maxima suppression, it is right
Background pixel point other than crack area is inhibited, then obtains the threshold value of weeds in the picture in proportion, if the pole of weeds
Big value is less than given threshold value and is finally come out endpoint detections with Canny then just will be deleted weeds.
5. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (7)
In, the generation method of three-dimensional point cloud is the principle of triangulation using similar triangles, between the depth of certain point and its parallax
There are following relationships:Wherein z is the depth distance of certain point, and d is parallax, and f is focal length, and b is between two cameras
Parallax range.
6. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (8)
In, for crackle, its three-dimensional point cloud is fitted using polynomial curve, to solve length, and width information then passes through meter
The distance for calculating both sides of the edge point in three-dimensional point cloud obtains；For pit, the boundary of its three-dimensional point cloud is extracted first, then using most
Boundary point is carried out ellipse fitting by small square law, obtains length of the elliptical long side as pit, elliptical short side is as pit
Width, the depth of pit then passes through three-dimensional point cloud pits and obtains at a distance from dam surface.
7. reservoir dam slope protection surface damage visible detection method according to claim 1, it is characterised in that: step (8)
In, a set of degree of injury judgment basis first is established to crackle and pit respectively: for crackle, using its length and width as foundation,
Impairment scale, minimum, the long and thick crackle grade highest of short and thin crackle grade are divided step by step according to length and width size；
For pit, using its size and the depth as foundation, impairment scale, small and shallow pit grade are divided according to size and the depth step by step
Impairment scale is minimum, and big and deep Notch damage grade is maximum；After obtaining the dimension information of crackle or pit, with the damage in backstage
Hurt grade judgment basis to compare, so that it is determined that degree of injury.
8. a kind of bionic device of reservoir dam slope protection surface damage vision-based detection, it is characterised in that: including bionical cement tree and
Binocular Stereo Vision System i.e. two microcam, using reservoir dam slope protection table according to any one of claims 1 to 7
Surface damage visible detection method；The middle part of bionical cement tree is provided with bionical egression head, and Binocular Stereo Vision System is mounted on imitative
On raw egression head, bionical egression head can adjust the angle of Binocular Stereo Vision System by rotating, and real-time monitoring reservoir is big
The damage of the slope protections such as the crackle or pit on dam slope protection surface.
9. the bionic device of reservoir dam slope protection surface damage vision-based detection according to claim 8, it is characterised in that: imitative
Unboiled water mud tree is to imitate coastal Anti-Typhoon cocoanut tree structure, and the cross section of bionical cement tree is approximate ellipsoidal, is provided with annual ring
Groove；The top of bionical cement tree is equipped with artificial tree top cover, and the periphery of artificial tree top cover is fixed with by retention mechanism
Solar panels and its battery；Solar panels are connected by electric wire with microcam, provide the energy for microcam；The knot
Structure can evacuate power and reduce resistance and vibration；Microcam is connect by cable and its transmitter with computer, will be miniature
The data transmission of video camera is to computer.
10. the bionic device of reservoir dam slope protection surface damage vision-based detection according to claim 8, it is characterised in that:
Bionical egression head is installed on bionical cement tree by screw, and the angle of microcam is adjusted by adjusting screw
Degree；It is provided with camera mounting bracket on bionical egression head, for installing microcam；The top of bionical egression head is provided with bionical
Leaf is lubricious material, for being microcam rain cover and sunlight.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910235217.3A CN110009610A (en) | 2019-03-27 | 2019-03-27 | Visual detection method for surface damage of reservoir dam protection slope and bionic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910235217.3A CN110009610A (en) | 2019-03-27 | 2019-03-27 | Visual detection method for surface damage of reservoir dam protection slope and bionic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110009610A true CN110009610A (en) | 2019-07-12 |
Family
ID=67168296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910235217.3A Pending CN110009610A (en) | 2019-03-27 | 2019-03-27 | Visual detection method for surface damage of reservoir dam protection slope and bionic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110009610A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111581709A (en) * | 2020-05-18 | 2020-08-25 | 中国水利水电科学研究院 | Earth-rock dam structure modeling method and device based on point cloud |
| CN111815694A (en) * | 2020-09-14 | 2020-10-23 | 湖南大学 | Fatigue crack propagation life prediction method, device, equipment and storage medium |
| CN111896550A (en) * | 2020-03-31 | 2020-11-06 | 广西师范大学 | Surface defect detection device and method |
| CN112529880A (en) * | 2020-12-16 | 2021-03-19 | 武汉巨合科技有限公司 | Dam image crack detection device and method based on binocular vision |
| CN113109339A (en) * | 2021-04-08 | 2021-07-13 | 仲恺农业工程学院 | Control method of defect visual inspection equipment |
| CN113343554A (en) * | 2021-04-15 | 2021-09-03 | 长江勘测规划设计研究有限责任公司 | Arch dam underwater damage identification method, terminal equipment and storage medium |
| CN113359573A (en) * | 2021-06-30 | 2021-09-07 | 华能澜沧江水电股份有限公司 | Intelligent dam safety prediction method and device |
| CN113870326A (en) * | 2021-09-01 | 2021-12-31 | 浙江大学 | Structural damage mapping, quantifying and visualizing method based on image and three-dimensional point cloud registration |
| CN114841995A (en) * | 2022-06-02 | 2022-08-02 | 西南交通大学 | Deep learning-based railway roadbed fender equipment service state evaluation method |
| CN117745713A (en) * | 2024-01-24 | 2024-03-22 | 广东省建筑工程监理有限公司 | Slope protection structure deformation detection method and system based on image processing |
| CN118071158A (en) * | 2024-04-22 | 2024-05-24 | 中铁水利信息科技有限公司 | Intelligent hydraulic engineering early warning system based on intelligent sensing and data fusion |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093113A (en) * | 2013-02-04 | 2013-05-08 | 河海大学 | Concrete dam damage field harmfulness diagnostic method |
| CN105160652A (en) * | 2015-07-10 | 2015-12-16 | 天津大学 | Handset casing testing apparatus and method based on computer vision |
| CN105894499A (en) * | 2016-03-25 | 2016-08-24 | 华南理工大学 | Binocular-vision-based rapid detection method for three-dimensional information of space object |
| CN107179322A (en) * | 2017-06-15 | 2017-09-19 | 长安大学 | A kind of bridge bottom crack detection method based on binocular vision |
| CN108195932A (en) * | 2017-11-30 | 2018-06-22 | 北京宇航系统工程研究所 | A kind of aircraft pipeline damages supersonic guide-wave quantitative evaluating method |
| WO2019050417A1 (en) * | 2017-09-06 | 2019-03-14 | Auckland Uniservices Limited | Stereoscopic system calibration and method |
-
2019
- 2019-03-27 CN CN201910235217.3A patent/CN110009610A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093113A (en) * | 2013-02-04 | 2013-05-08 | 河海大学 | Concrete dam damage field harmfulness diagnostic method |
| CN105160652A (en) * | 2015-07-10 | 2015-12-16 | 天津大学 | Handset casing testing apparatus and method based on computer vision |
| CN105894499A (en) * | 2016-03-25 | 2016-08-24 | 华南理工大学 | Binocular-vision-based rapid detection method for three-dimensional information of space object |
| CN107179322A (en) * | 2017-06-15 | 2017-09-19 | 长安大学 | A kind of bridge bottom crack detection method based on binocular vision |
| WO2019050417A1 (en) * | 2017-09-06 | 2019-03-14 | Auckland Uniservices Limited | Stereoscopic system calibration and method |
| CN108195932A (en) * | 2017-11-30 | 2018-06-22 | 北京宇航系统工程研究所 | A kind of aircraft pipeline damages supersonic guide-wave quantitative evaluating method |
Non-Patent Citations (2)
| Title |
|---|
| 于海川: "《自适应Canny算法在钢板缺陷边缘检测中的应用》", 《软件导刊》, pages 175 - 177 * |
| 赵静: "《融合Canny算子和小波变换的边缘检测方法》", 《计算机仿真》, pages 277 - 280 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111896550A (en) * | 2020-03-31 | 2020-11-06 | 广西师范大学 | Surface defect detection device and method |
| CN111581709B (en) * | 2020-05-18 | 2021-03-16 | 中国水利水电科学研究院 | Earth-rock dam structure modeling method and device based on point cloud |
| CN111581709A (en) * | 2020-05-18 | 2020-08-25 | 中国水利水电科学研究院 | Earth-rock dam structure modeling method and device based on point cloud |
| CN111815694A (en) * | 2020-09-14 | 2020-10-23 | 湖南大学 | Fatigue crack propagation life prediction method, device, equipment and storage medium |
| CN112529880A (en) * | 2020-12-16 | 2021-03-19 | 武汉巨合科技有限公司 | Dam image crack detection device and method based on binocular vision |
| CN113109339A (en) * | 2021-04-08 | 2021-07-13 | 仲恺农业工程学院 | Control method of defect visual inspection equipment |
| CN113343554B (en) * | 2021-04-15 | 2023-04-25 | 长江勘测规划设计研究有限责任公司 | Arch dam underwater damage identification method, terminal equipment and storage medium |
| CN113343554A (en) * | 2021-04-15 | 2021-09-03 | 长江勘测规划设计研究有限责任公司 | Arch dam underwater damage identification method, terminal equipment and storage medium |
| CN113359573A (en) * | 2021-06-30 | 2021-09-07 | 华能澜沧江水电股份有限公司 | Intelligent dam safety prediction method and device |
| CN113870326A (en) * | 2021-09-01 | 2021-12-31 | 浙江大学 | Structural damage mapping, quantifying and visualizing method based on image and three-dimensional point cloud registration |
| CN113870326B (en) * | 2021-09-01 | 2023-11-28 | 浙江大学 | Structural damage mapping, quantifying and visualizing method based on image and three-dimensional point cloud registration |
| CN114841995A (en) * | 2022-06-02 | 2022-08-02 | 西南交通大学 | Deep learning-based railway roadbed fender equipment service state evaluation method |
| CN117745713A (en) * | 2024-01-24 | 2024-03-22 | 广东省建筑工程监理有限公司 | Slope protection structure deformation detection method and system based on image processing |
| CN117745713B (en) * | 2024-01-24 | 2024-07-23 | 广东省建筑工程监理有限公司 | Slope protection structure deformation detection method and system based on image processing |
| CN118071158A (en) * | 2024-04-22 | 2024-05-24 | 中铁水利信息科技有限公司 | Intelligent hydraulic engineering early warning system based on intelligent sensing and data fusion |
| CN118071158B (en) * | 2024-04-22 | 2024-07-05 | 中铁水利信息科技有限公司 | Intelligent hydraulic engineering early warning system based on intelligent sensing and data fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110009610A (en) | Visual detection method for surface damage of reservoir dam protection slope and bionic device | |
| US11486548B2 (en) | System for detecting crack growth of asphalt pavement based on binocular image analysis | |
| CN111855664B (en) | Adjustable three-dimensional tunnel defect detection system | |
| CN102175613B (en) | Image-brightness-characteristic-based pan/tilt/zoom (PTZ) video visibility detection method | |
| CN105225482B (en) | Vehicle detecting system and method based on binocular stereo vision | |
| CN104484648B (en) | Robot variable visual angle obstacle detection method based on outline identification | |
| CN101216304B (en) | Systems and methods for object dimension estimation | |
| CN104408746B (en) | A kind of passenger flow statistical system based on depth information | |
| CN106156758B (en) | A kind of tidal saltmarsh method in SAR seashore image | |
| CN110031476A (en) | Multi-view stereoscopic vision detection device and method for loosening or breaking of nut and screw of bridge base | |
| CN209708175U (en) | Reservoir dam bank protection surface damage visual inspection bionic device | |
| CN104182992B (en) | Method for detecting small targets on the sea on the basis of panoramic vision | |
| CN110675449A (en) | Binocular camera-based offshore flow detection method | |
| CN111476762B (en) | Obstacle detection method and device of inspection equipment and inspection equipment | |
| CN106774439A (en) | Solar tracking bearing calibration and device based on solar motion rule and IMAQ | |
| CN104330048A (en) | Image-based railway snow depth measurement device and method | |
| CN116545122B (en) | Power transmission line external damage prevention monitoring device and external damage prevention monitoring method | |
| CN115060343A (en) | Point cloud-based river water level detection system, detection method and program product | |
| CN116110081A (en) | Detection method and system for wearing safety helmet based on deep learning | |
| CN115620239A (en) | Point cloud and video combined power transmission line online monitoring method and system | |
| CN209230716U (en) | A kind of volume measurement device | |
| CN114067533A (en) | Geological disaster photographing monitoring and early warning method | |
| CN104573662A (en) | Cloud judgment method and system | |
| CN213874180U (en) | Binocular vision measurement system for thermal power plant pipeline expansion amount | |
| CN106709432A (en) | Binocular stereoscopic vision based head detecting and counting method |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190712 |
|
| WD01 | Invention patent application deemed withdrawn after publication |