CN110307790A - Camera shooting machine detecting device and method applied to safety monitoring slope - Google Patents
Camera shooting machine detecting device and method applied to safety monitoring slope Download PDFInfo
- Publication number
- CN110307790A CN110307790A CN201910599471.1A CN201910599471A CN110307790A CN 110307790 A CN110307790 A CN 110307790A CN 201910599471 A CN201910599471 A CN 201910599471A CN 110307790 A CN110307790 A CN 110307790A
- Authority
- CN
- China
- Prior art keywords
- point
- video camera
- slope
- image
- monitoring
- 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
Classifications
-
- 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
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
- G01B11/03—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness by measuring coordinates of points
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C11/00—Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
- G01C11/04—Interpretation of pictures
- G01C11/06—Interpretation of pictures by comparison of two or more pictures of the same area
- G01C11/08—Interpretation of pictures by comparison of two or more pictures of the same area the pictures not being supported in the same relative position as when they were taken
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of video camera detection method and device applied to safety monitoring slope, for the complexity of side slope, readily identified identifying body is set in side slope, reflect the state of side slope by these identifying body, its core is to obtain the space coordinate of target array, on side slope surface, a certain number of monitoring objective points are set, side slope image is acquired with 2 edge of table slope surface displacement intelligent high-definition video cameras, the image coordinate of monitoring point is extracted using sub-pix searching algorithm, pass through photo coordinate system, camera coordinate system, measure the transformational relation between coordinate system, calculate the three-dimensional coordinate of monitoring point, identify side slope surface displacement.Therefore, the scheme of the application has the advantage that feasibility is strong, is easily achieved.
Description
Technical field
The present invention relates to technical field of image processing, specifically a kind of camera shooting machine testing applied to safety monitoring slope is filled
It sets and method.
Background technique
Since the 1990s, as economic construction of China develops, the requirement to highway communication is also higher and higher.By
In China's landforms, the restriction of geological conditions limitation and highway alignment, slope problem caused by height is filled out, deep-cut is very universal.One
Slope hazard occurs for denier, and consequence is extremely serious.Deformation is the most significant parameter for characterizing side slope variation, best embodies the variation of side slope
Situation and development trend.If can accurately detect the displacement of side slope and be forecast to the development trend of deformation, so that it may
To take measures on customs clearance as early as possible and countermeasure, it is preferably minimized loss caused by disaster.
It is at present deformation observation for the common monitoring method of slope monitoring: surveys the displacement, sedimentation and deformation feelings of side slope
Condition；Stress trajectory: monitoring edge slope structure object, the safety of safeguard structure and protection effect；And porous flow pattern: side slope is monitored
Seepage flow and groundwater change.Key instrument has traditional inclinometer pipe, pressure gauge, displacement meter etc., due to monitoring instrument and computer
The fast development of technology has also appeared a variety of emerging monitoring technology, such as novel GPS, optical sensor, TDR.The above method
It is the devices such as embedding strain or sensing inside side slope, the stabilization of side slope is detected by supplementary means such as relative computer softwares
Property, all there is the disadvantages of highly specialized, at high cost, risk is big to varying degrees.
For this purpose, Authorization Notice No. is the patent of invention of CN106323176B, the three of a kind of open-pit slope are disclosed
Displacement monitoring method is tieed up, this method is the point cloud data that opencut side slope region to be monitored is obtained by three-dimensional laser scanner,
Establish side slope three-dimensional model to be monitored；The monitoring imaged image for acquiring side slope region to be monitored in real time by monocular-camera, is obtained
Take slope monitoring sequential images image；Control is chosen in the side slope regional scope to be monitored of slope monitoring sequential images image covering
It is processed, determine direct linear transformation's equation that geodetic coordinates and imaged image pixel two-dimensional coordinate are mutually converted；Using image
Local feature detection and matching algorithm, extract the characteristic point to match in real-time monitoring slope monitoring sequential images image, real
When monitoring slope monitoring sequential images image in the position of characteristic point that matches in current image image and upper width imaged image
It moves；The two dimensional image displacement of characteristic point in slope monitoring sequential images image is converted to the three of earth coordinates slope point
Dimension displacement.However in the patent, the acquisition of cloud is put, generally passes through three-dimensional imaging sensor, such as binocular camera, 3-D scanning
Instrument, RGB-D camera etc., which adds the costs of equipment.Point cloud first is not dense expression, generally than sparse,
It is exaggerated and sees, can be appreciated that point isolated one by one.It is black around it, that is, without information.So many in space
A cloud is not put in position in fact, what the information of this part was missing from.Obvious characteristic part as just lacked identifying body cannot
Reach the desired effect of slope test.Next puts the resolution ratio of cloud and the distance dependent from camera, closely sees that identification point very may be used
It can not see, the far three-dimensional slope model and a practical feelings that can just see one probably, pass through that point cloud is established in this way that can only be drawn
Condition just easily causes error.In addition, the patent needs first to establish three-dimensional side after extracting identification point parameter to real time monitoring picture
Threedimensional model, either computational efficiency or standard are substituted into after extracting characteristic point by the picture of monocular-camera shooting again after the model of slope
Very big influence of the exactness by three-dimensional slope model.
Summary of the invention
The brief overview about the embodiment of the present invention is given below, in order to provide about certain aspects of the invention
Basic comprehension.It should be appreciated that outlined below is not about exhaustive general introduction of the invention.It is not intended to determine this hair
Bright key or pith, nor is it intended to limit the scope of the present invention.Its purpose only provides certain in simplified form
A little concepts, taking this as a prelude to a more detailed description discussed later.
According to the one aspect of the application, a kind of video camera detection method applied to safety monitoring slope, the party are provided
Method includes:
Step 1: establishing object space coordinate system, which describes the space coordinate where target, gets mesh
Standard type parameter, and it is stored in database；Wherein, object space coordinate system also referred to as measures coordinate system, is a reference frame, is used for
The absolute position for describing object and video camera, is the absolute coordinate of objective world；
Step 2: a video camera being at least respectively set at left and right sides of side slope, is denoted as left video camera and right video camera；It is left
Video camera and right video camera acquire the slope monitoring image in side slope region to be monitored from least two different angles；Image can be
The image acquired at regular intervals is also possible to the video acquired in real time；
Step 3: image is monitored by left and right cameras side slope and carries out Feature Points Matching, and is identified the identification of body,
Extract identifying body parameter；Identifying body is the preset identification module based on color, shapes and sizes parameter, such as identifying body
It can be the square module that vertical angles are identified；
Step 4: identifying body parameter and objective body parameter being compared, if do not changed, show that position does not occur
It moves；Continue monitoring until changing, then carries out three-dimensional computations, obtain slope displacement variable quantity.
Wherein, image progress Feature Points Matching is monitored by left and right cameras side slope to comprise the following processes:
Process A1: the characteristic point coordinate geometry of left images, respectively L and R are inputted；
Process A2: traversal set L calculates each characteristic point and corresponds to polar curve equation in right image, then each point on calculating R
To the distance of every polar curve equation, if distance is less than 1 pixel, then it is assumed that the point is candidate point；
Process A3: if candidate point number is 1, left figure picture point will be changed and be put into set Lt, candidate point is put into set Rt, Lt
And RtFor the set of characteristic points of successful match.If candidate point number is greater than 1, left figure picture point will be changed and be put into set Lh, candidate point
Set is put into set Rh。
Process A4: traversal set Lh, to each characteristic pointIn LtIn take Euclidean distance nearest characteristic point as left detent
PointCorresponding RtIn characteristic point be right detent point
Process A5: left detent point is calculatedWith characteristic pointVector angleAnd Euclidean distance
Process A6: traversal characteristic pointCorresponding RhIn candidate point set, calculate right detent pointWith each candidate point
Vector angleAnd Euclidean distanceAnd withWithIt makes the difference.Setting minimum angles threshold values 10, minimum range threshold values 10,
The candidate point for finding difference minimum and branch threshold requirements of waiting a moment candidate o'clock is match point, and left figure characteristic point is put into Lt, right
Figure match point is put into Rt.Until having traversed Lh, can be obtained and left figure characteristic point matched right figure characteristic point one by one.
In step 3, the identification of identifying body is comprised the following processes:
Process B1: left and right cameras carries out Feature Points Matching, gets the image containing identifying body, and progress color first is sentenced
Other: get colors bright-coloured identifying body, is different from the environment of surrounding；HIS color segmentation can be used to face specific in color image
It is effectively extracted in color region.
Process B2: and then carry out shape discrimination: binaryzation, edge detection, contours extract, rectangular inspection are carried out to identifying body
It surveys: being specifically that gaussian filtering operation first is carried out to image, eliminate the Gaussian noise in image, binaryzation behaviour then is carried out to image
Make, tentatively obtains square region block.Edge detection is carried out to the region unit in binary picture by Canny edge detection algorithm,
Obtain preliminary square rim.Contours extract be in order to extract each independent square block, also needed after the completion of extracting by
Noise remove step removes the excessive or too small profile of too large or too small profile and Aspect Ratio, can obtain from people is obtained
To more meeting rectangular profile.
Process B3: and then size differentiation is carried out using model matching method, because this two after color and shape differentiates
A parameter has become known parameters, sufficiently designs a model equal with identifying body projected size with known parameters, counts
Calculate similarity, similarity it is high be determined as target object.
1-3 through the above steps constructs three-dimensional scenic, is detecting using the parallax of the two images of twin-lens intake
After moving target, by calculating the position deviation between image corresponding points, the three-dimensional information of target is obtained, to mesh in depth image
High-precision displacement judgement is realized in target detection and tracking.I.e. at the center for detecting identifying body (positioning identifier, observation mark)
After being worth coordinate, need to establish a set of Displacement predication standard to judge whether identifying body is displaced.Specifically: being in side slope
All observation is calculated when stable state respectively and is identified to the distance of positioning identifier, and is stored in database.In detection later
In, one " warning value " is set, determines whether side slope is displaced by comparing whether linear distance is more than " warning value ",
And specific direction of displacement and actual displacement amount are obtained by three-dimensional resolving.Video camera can be effectively overcome with this method
Error caused by shaking.
It is matched by features described above point and identifying body identification can be obtained whether side slope is displaced, next only needed
Calculated by Three-Dimensional Solution and calculates direction of displacement and actual displacement amount.Specifically, three-dimensional computations are particular by analysis as plane is sat
System, camera coordinate system, the transformational relation between object space coordinate system (measurement coordinate system) are marked, the three-dimensional seat of monitoring point is calculated
Mark.This method can accurately detect side slope surface displacement and can predict the trend of deformation, so as to adopt as early as possible
Take relevant countermeasure and measure.Specifically, the calculating process of the three-dimensional coordinate of the calculating monitoring point of step 4 includes:
Process 1: establishing photo coordinate system O1xy, indicates are as follows:
In formula: ku, kv are respectively length factor of the unit pixel in x-axis, y-axis direction；(Uo, Vo) is photo coordinate system
The image coordinate of original place O1；
Establish camera coordinate system OcXcYcZc；
It establishes object space coordinate system: describing object space coordinate system and camera coordinates with translation vector t and spin matrix R
Relationship between system, if homogeneous coordinates of the point P under camera coordinate system in space are (Xc, Yc, Zc) T, the two exists
Following relationship:
Wherein t is 3 dimension translation vectors, and R is 3*4 unit matrix, Ot=(0,0,0)T, M is relational matrix, is two coordinates
Tie between system.
Process 2: the three-dimensional coordinate of monitoring point calculates:
Relationship between the image coordinate and object space coordinate system of monitoring point is expressed as:
Assuming that M is the parameter matrix of 3*4 rank, may be expressed as:
If two camera parameters matrixes are respectively MA and MB, can obtain:
(ma 11-uama 31)Xw+(ma 12-uama 32)Yw+(ma 13-uama 33) Zw=(uama 34-ma 14)；
(ma 21-vama 31)Xw+(ma 22-vama 32)Yw+(ma 23-vama 33) Zw=(vama 34-ma 24)；
(mb 11-ubma 31)Xw+(mb 12-ubmb 32)Yw+(mb 13-ubmb 33) Zw=(ubmb 34-mb 14)；
(mb 21-vbmb 31)Xw+(mb 22-vbmb 32)Yw+(mb 23-vbmb 33) Zw=(vbmb 34-mb 24)；
Using least square method solving equations, the object space three-dimensional coordinate (x, y, z) of monitoring objective point can be calculated；
After finding out the three-dimensional coordinate of each observation mark, it is compared with the coordinate under original state respectively, i.e.,
Obtain the actual displacement amount that side slope surface occurs.
In image measurement process and machine vision applications, for the three-dimensional geometry position for determining space object surface point
With the correlation between its in the picture corresponding points, it is necessary to establish the geometrical model of video camera imaging.Video camera in the application
Calibration be unusual the key link, the precision of calibration result and the stability of algorithm directly affect camera operation and generate knot
The accuracy of fruit.Further, before step 2 acquires image, there are also the processes demarcated to video camera:
Observation mark is directly fabricated to scaling board M, can directly determine homography matrix a H, an orthogonal matrix R.H
H column vector is unfolded, is had by=sMR:
H=[h1 h2h3]=sM [r1 r2t]；
Solution obtains:
Because R is orthogonal matrix, r1 and r2 are unit orthogonal vectors, i.e., its inner product is 0, mould 1.
Then have
Acquire following formula:
h1M-TM-1h2=0；
If B=M-TM-1, M is the matrix of 3*3, i.e. B is a symmetrical matrix, can be unfolded are as follows:
Because B is symmetrical matrix, only 6 times are original, Matrix Multiplication opened available:
Available N number of identifying body image simultaneously, stacking these equations has: Vb=0；
Wherein V is the matrix of 2N*6, and as K > 2, b has solution.Camera intrinsic parameter can be from the closing Xie Zhongzhi of B matrix
It connects to obtain fx、fy、cx、cyValue.
It has been calculated before outer parameter:
r3=r1*r2；
Acquire the calibration that video camera is completed after the inside and outside parameter of video camera.
According to further aspect of the application, a kind of camera shooting machine detecting device applied to safety monitoring slope is provided, is wrapped
It includes:
Object space coordinate system establishes unit, which describes the space coordinate where target, gets target
Body parameter, and it is stored in database；
Slope monitoring image capturing unit, is at least respectively set a video camera at left and right sides of side slope, is denoted as left camera shooting
Machine and right video camera；Left video camera and right video camera acquire the slope monitoring in side slope region to be monitored from least two different angles
Image；
Identifying body parameter extraction unit carries out Feature Points Matching by left and right cameras, and is identified the identification of body, mentions
Take identifying body parameter；Identifying body is the preset identification module based on color, shapes and sizes parameter；
Slope displacement variable quantity monitoring unit, identifying body parameter and objective body parameter are compared, if do not become
Change, then shows not to be subjected to displacement；Continue monitoring until changing, then carries out three-dimensional computations, obtain slope displacement variable quantity.
The scheme of existing detection slope stability requires the devices such as embedding strain or sensing inside side slope, is this this Shen
The side slope surface displacement intelligent high-definition video camera detection device based on image recognition please be propose, is carried out using binocular vision technology
Image recognition, it is different from traditional " point " measurement monitoring method, is a kind of non-contact technology based on " face " measurement, has energy
The advantages of reaching rapid survey structure surface three dimension coordinate, and realizing remote network monitoring.
Camera shooting machine detecting device of the invention, image detection are reached using image procossing and identification technology to target
Parameter obtains, and has the characteristics that limit by range, non-contact and dynamic reflection slope deforming overall picture, the machine with routine
The incomparable advantage of the methods of tool, electrical measurement, flash ranging.And it is directed to the complexity of side slope, the application is arranged in side slope and is easy to
The identifying body of identification reflects the state of side slope by these identifying body, and core is to obtain the space coordinate of target array, on side
A certain number of monitoring objective points are arranged in slope surface, acquire side slope image with 2 edge of table slope surface displacement intelligent high-definition video cameras, adopt
The image coordinate that monitoring point is extracted with sub-pix searching algorithm passes through photo coordinate system, camera coordinate system, measurement coordinate system
Between transformational relation, calculate the three-dimensional coordinate of monitoring point, identify side slope surface displacement.Therefore, the scheme of the application have can
The advantage that row is strong, is easily achieved.
Detailed description of the invention
The present invention can be by reference to being better understood, wherein in institute below in association with description given by attached drawing
Have and has used the same or similar appended drawing reference in attached drawing to indicate same or similar component.The attached drawing is together with following
It is described in detail together comprising in the present specification and forming a part of this specification, and is used to that this is further illustrated
The preferred embodiment and explanation the principle of the present invention and advantage of invention.In the accompanying drawings:
Fig. 1 is the building-block of logic of camera shooting machine detecting device of the invention；
Fig. 2 is the objective body and object space coordinate diagram of camera shooting machine detecting device of the invention；
Fig. 3 is the schematic diagram of plane coordinate system of the invention, camera coordinate system, object space coordinate system；
Fig. 4 is the distortion schematic diagram of camera shooting machine detecting device of the invention.
Specific embodiment
Illustrate the embodiment of the present invention below with reference to accompanying drawings.For purposes of clarity, it is omitted in attached drawing and explanation
Unrelated to the invention, component known to persons of ordinary skill in the art and processing expression and description.
Referring to Fig. 1, the camera shooting machine detecting device applied to safety monitoring slope of the invention realizes that slope test is divided into three
A part, first part is headend equipment, including monitoring point is arranged in side slope, arranges more industrial face battle array colour TV cameras,
Object space coordinate system is established, calibration point coordinate is obtained；Second step point is system front end, is examined by the comparison of identifying body and objective body
Survey whether slope displacement changes in image level, if there is no variation, it is lasting to monitor；After Part III is system
Platform carries out three-dimensional resolving, calculates slope displacement actual displacement variable quantity if there is a change.With traditional displacement detecting method
It compares, the advantage of the present invention with high security, at low cost, rapid, reliable.
As a specific example, camera shooting machine detecting device of the invention includes: multiple monitoring objective points 1, on composition
The monitoring point layout area in 2 region of side slope；The monitoring point layout area in 3 region of slope similarly obtains.
Specifically, the video camera detection method applied to safety monitoring slope of the invention, this method comprises:
Step 1: establishing object space coordinate system, which describes the space coordinate where target, gets mesh
Standard type parameter, and it is stored in database；Wherein, object space coordinate system also referred to as measures coordinate system, uses rectangular coordinate system in space
OwZwYwZw is indicated, is a reference frame, is the absolute of objective world for describing the absolute position of object and video camera
Coordinate；
Step 2: a video camera being at least respectively set at left and right sides of side slope, is denoted as left video camera and right video camera；It is left
Video camera and right video camera acquire the slope monitoring image in side slope region to be monitored from least two different angles；Image can be
The image acquired at regular intervals is also possible to the video acquired in real time；
Step 3: image is monitored by left and right cameras side slope and carries out Feature Points Matching, and is identified the identification of body,
Extract identifying body parameter；In the present embodiment, identifying body can be the square module that vertical angles are identified；
Step 4: identifying body parameter and objective body parameter being compared, if do not changed, show that position does not occur
It moves；Lasting monitoring, if there is a change, then carries out three-dimensional computations, obtains slope displacement variable quantity.
In the present embodiment, referring to fig. 2, two video cameras are arranged in a monitoring point, and it is colored that industrial face battle array can be selected in video camera
Video camera.Video camera shoots side slope video, on the one hand by playing after coding and decoding in system background, on the one hand extracts side slope figure
Picture is identified the identification of body, extracts identifying body parameter.Obtained identifying body parameter and objective body parameter are compared,
It is lasting to monitor if do not changed, backstage is passed to while alarm if there is a change and carries out three-dimensional computations, obtains side slope position
Move variable quantity.To which accurate measurements go out the displacement of side slope and forecast to the development trend of deformation, so that it may take phase as early as possible
Pass measure and countermeasure are preferably minimized loss caused by disaster.
Wherein, due to the complexity of side slope, the side slope statement model of abstract universality is difficult, therefore, the application wound
Readily identified identifying body is arranged in the property made in side slope, and the state of side slope is reflected by these identifying body.Meanwhile in reality
In processing, the two dimension slope monitoring image of the only left video camera obtained in computer and the acquisition of right video camera, it is therefore desirable to right
The image of left and right cameras carries out Feature Points Matching.For the weakness of traditional Stereo matching, the Feature Points Matching of the application
Under epipolar-line constraint existing for binocular camera itself, increase angle restriction.Conventional Stereo matching is exactly in two images M1
With the point for finding same characteristic features on M2, it is necessary first to it is determined a bit on M1, corresponding match point is then found on M2, if
Not certain constraint condition, the process of found match point are gamuts, this is a kind of very time-consuming operation, and is sought
It looks for the probability of match point mistake also can be very big, is not wise selection.There are certain relationships, i.e. pole for binocular vision model itself
The searching of match point can be tied on straight line by line constraint, epipolar-line constraint, substantially increase progress in this way, also greatly
Reduce match time.In practical Stereo matching, although epipolar-line constraint greatly reduces match time, but working as has multiple spies
Sign point when, still can there is a certain error, to solve this problem, this patent propose it is a kind of increase angle restriction characteristic point
It with algorithm, makes full use of without dispute match point angle, matches the controversial characteristic point under epipolar-line constraint, specifically, step 3 is logical
Left and right cameras side slope monitoring image progress Feature Points Matching is crossed to comprise the following processes:
Process A1: the characteristic point coordinate geometry of left images, respectively L and R are inputted；
Process A2: traversal set L calculates each characteristic point and corresponds to polar curve equation in right image, then each point on calculating R
To the distance of every polar curve equation, if distance is less than 1 pixel, then it is assumed that the point is candidate point；
Process A3: if candidate point number is 1, left figure picture point will be changed and be put into set Lt, candidate point is put into set Rt, Lt
And RtFor the set of characteristic points of successful match.If candidate point number is greater than 1, left figure picture point will be changed and be put into set Lh, candidate point
Set is put into set Rh。
Process A4: traversal set Lh, to each characteristic pointIn LtIn take Euclidean distance nearest characteristic point as left detent
PointCorresponding RtIn characteristic point be right detent point
Process A5: left detent point is calculatedWith characteristic pointVector angleAnd Euclidean distance
Process A6: traversal characteristic pointCorresponding RhIn candidate point set, calculate right detent pointWith each candidate point
Vector angleAnd Euclidean distanceAnd withWithIt makes the difference.Setting minimum angles threshold values 10, minimum range threshold values 10,
The candidate point for finding difference minimum and branch threshold requirements of waiting a moment candidate o'clock is match point, and left figure characteristic point is put into Lt, right
Figure match point is put into Rt.Until having traversed Lh, can be obtained and left figure characteristic point matched right figure characteristic point one by one.
In step 3, identifying body is divided into positioning identifier and observation mark, the face of identifying body being identified by identifying body
Color, shape, size are differentiated that final identification marking body specifically includes following process:
Process B1, left and right cameras carry out Feature Points Matching, get the image containing identifying body, and progress color first is sentenced
Other: get colors bright-coloured identifying body, is different from the environment of surrounding；HIS color segmentation can be used to face specific in color image
It is effectively extracted in color region.
Process B2: and then carry out shape discrimination: binaryzation, edge detection, contours extract, rectangular inspection are carried out to identifying body
It surveys: being specifically that gaussian filtering operation first is carried out to image, eliminate the Gaussian noise in image, binaryzation behaviour then is carried out to image
Make, tentatively obtains square region block.Edge detection is carried out to the region unit in binary picture by Canny edge detection algorithm,
Obtain preliminary square rim.Contours extract be in order to extract each independent square block, also needed after the completion of extracting by
Noise remove step removes the excessive or too small profile of too large or too small profile and Aspect Ratio, can obtain from people is obtained
To more meeting rectangular profile.
Process B3: and then size differentiation is carried out using model matching method, because this two after color and shape differentiates
A parameter has become known parameters, sufficiently designs a model equal with identifying body projected size with known parameters, counts
Calculate similarity, similarity it is high be determined as target object.
1-3 through the above steps constructs three-dimensional scenic, is detecting using the parallax of the two images of twin-lens intake
After moving target, by calculating the position deviation between image corresponding points, the three-dimensional information of target is obtained, to mesh in depth image
High-precision displacement judgement is realized in target detection and tracking.I.e. at the center for detecting identifying body (positioning identifier, observation mark)
After being worth coordinate, need to establish a set of Displacement predication standard to judge whether identifying body is displaced.Specifically: being in side slope
All observation is calculated when stable state respectively and is identified to the distance of positioning identifier, and is stored in database.In detection later
In, one " warning value " is set, determines whether side slope is displaced by comparing whether linear distance is more than " warning value ",
And specific direction of displacement and actual displacement amount are obtained by three-dimensional resolving.Video camera can be effectively overcome with this method
Error caused by shaking.
It is matched by features described above point and identifying body identification can be obtained whether side slope is displaced, next only needed
Calculated by Three-Dimensional Solution and calculates direction of displacement and actual displacement amount.Specifically, three-dimensional computations are particular by analysis as plane is sat
System, camera coordinate system, the transformational relation between object space coordinate system (measurement coordinate system) are marked, the three-dimensional seat of monitoring point is calculated
Mark.This method can accurately detect side slope surface displacement and can predict the trend of deformation, so as to adopt as early as possible
Take relevant countermeasure and measure.Specifically, the calculating process of the three-dimensional coordinate of the calculating monitoring point of step 4 includes:
Process 1: establishing photo coordinate system O1xy, indicates are as follows:
In formula: ku, kv are respectively length factor of the unit pixel in x-axis, y-axis direction；(Uo, Vo) is photo coordinate system
The image coordinate of original place O1；
Establish camera coordinate system OcXcYcZc；
It establishes object space coordinate system: describing object space coordinate system and camera coordinates with translation vector t and spin matrix R
Relationship between system, if homogeneous coordinates of the point P under camera coordinate system in space are (Xc, Yc, Zc) T, the two exists
Following relationship:
Wherein t is 3 dimension translation vectors, and R is 3*4 unit matrix, Ot=(0,0,0)T, M is relational matrix, is two coordinates
Tie between system.
Process 2: the three-dimensional coordinate of monitoring point calculates:
Relationship between the image coordinate and object space coordinate system of monitoring point is expressed as:
Assuming that M is the parameter matrix of 3*4 rank, may be expressed as:
If two camera parameters matrixes are respectively MA and MB, can obtain:
(ma 11—uama 31)Xw+(ma 12—uama 32)Yw+(ma 13—uama 33) Zw=(uama 34—ma 14)；
(ma 21—vama 31)Xw+(ma 22—vama 32)Yw+(ma 23—vama 33) Zw=(vama 34—ma 24)；
(mb 11—ubma 31)Xw+(mb 12—ubmb 32)Yw+(mb 13—ubmb 33) Zw=(ubmb 34—mb 14)；
(mb 21—vbmb 31)Xw+(mb 22—vbmb 32)Yw+(mb 23—vbmb 33) Zw=(vbmb 34—mb 24)；
Using least square method solving equations, the object space three-dimensional coordinate (x, y, z) of monitoring objective point can be calculated；
After finding out the three-dimensional coordinate of each observation mark, it is compared with the coordinate under original state respectively, i.e.,
Obtain the actual displacement amount that side slope surface occurs.
As shown in Fig. 2, objective body of the invention and object space coordinate system are established, hair is searched for using sub-pix angle point and obtains figure
As coordinate, theoretical precision can reach the 1/50 of pixel, i.e., if the physical length of pixel is 1mm, precision can reach
0.02mm。
In addition, the application is directed to characteristic of the video camera in application to slope, a kind of new thinking progress Three-Dimensional Solution is also designed
It calculates.After acquiring the parameters such as projection array target centroid, area, with the pin hole displacement battle array that is constituted of projection and projection array target centroid,
The spatial relationship of the parameters such as area calculates camera distance battle array, observes mark and positioning identifier projector distance battle array, depth difference battle array etc.
Distance parameter, all kinds of range informations needed for obtaining three-dimensional resolve.Specifically, this three resolvings comprise the following processes:
Observation is identified with according to pin hole projection:
Similarly, positioning identifier has:
WhereinProjection array area for observation mark i in t moment, i.e., the area identified in image,Observation mark
I is known in the real area of t moment.Indicate that t moment video camera identifies the distance of i to observation.It is positioning identifier 1 in t
The projection array area at moment.Distance of the video camera to positioning identifier 1.F is focal length of camera.
Simultaneous has:
Similarly, have for positioning identifier n:
Under projected coordinate system, B1、BnWith AiProjection is away from being respectively as follows:
The projector distance be exactly in the picture positioning identifier to realize two dimension to three-dimensional at a distance from observation mark
Conversion, it is also necessary to the value of depth difference.
Enable depth difference:
Raw footage is obtained by correction, then obtains correct depth value by projecting scaling relationship.
In known camera distance battle array, observation mark and object sky after positioning identifier projector distance battle array, depth difference battle array, can be solved
Between A under coordinate systemiThree-dimensional coordinate (x, y, z).
Side slope surface displacement S is calculated:
The starting material space coordinate of monitoring point is stored as (X on side slope surfacen0, Yn0, Zn0), the coordinate of t moment is (Xnt,
Ynt, Znt), then have
In formula: SnFor the displacement of n-th of measuring point；S is side slope surface displacement；N=1,2 ..., N, wherein N is total measuring point number.
In image measurement process and machine vision applications, for the three-dimensional geometry position for determining space object surface point
With the correlation between its in the picture corresponding points, it is necessary to establish the geometrical model of video camera imaging.Video camera in the application
Calibration be unusual the key link, the precision of calibration result and the stability of algorithm directly affect camera operation and generate knot
The accuracy of fruit.Further, before step 2 acquires image, there are also the processes demarcated to video camera:
Observation mark is directly fabricated to scaling board M, can directly determine homography matrix a H, an orthogonal matrix R.H
H column vector is unfolded, is had by=sMR:
H=[h1 h2 h3]=sM [r1 r2t]；
Solution obtains:
Because R is orthogonal matrix, r1 and r2 are unit orthogonal vectors, i.e., its inner product is 0, mould 1.
Then have
Acquire following formula:
h1M-TM-1h2=0；
If B=M-TM-1, M is the matrix of 3*3, i.e. B is a symmetrical matrix, can be unfolded are as follows:
Because B is symmetrical matrix, only 6 times are original, Matrix Multiplication opened available:
Available N number of identifying body image simultaneously, stacking these equations has: Vb=0；
Wherein V is the matrix of 2N*6, and as K > 2, b has solution.Camera intrinsic parameter can be from the closing Xie Zhongzhi of B matrix
It connects to obtain fx、fy、cx、cyValue.
It has been calculated before outer parameter:
r3=r1*r2；
Acquire the calibration that video camera is completed after the inside and outside parameter of video camera.Wherein, the application takes the photograph left and right using above-mentioned
The scheme that camera is demarcated is to improve realization on the basis of Zhang Zhengyou plane reference method, by its conventional chessboard
Scaling board replaces with the color, shape, size that front is chosen and specifically observes mark as scaling board.It, can be into using the program
One step improves the precision of calibration result and the stability of algorithm, and then improves the accuracy that camera operation generates result.Camera shooting
After machine calibration, the three-dimensional three-dimensional coordinate resolved to determine certain point in identifying body is carried out.
In computer calibration process, linear model is a kind of perfect condition, and under real conditions, due to the folding of camera lens
Penetrate error etc., imaging can bring different degrees of distortion, so that imaging is in the position (X, Y) of linear model description,
And the time coordinate deviated is (X1, Y1).Here only consider the radial distortion and tangential distortion being affected to calibration result.It enables
(xp, yp) it is the position really put, enable (xd, yd) it is distorted position, in conjunction with radial distortion and tangential distortion principle, can be obtained:
The application enormously simplifies current complex slope treatment process by the method for discrimination of special identifying body.Together
When, improvement also has been carried out to identifying body identification process, it can be achieved that the superior accuracy of conclusion.In addition, also to camera calibration
Improve, to adapt to the scheme of the application, and for the radial direction of traditional cameras calibration in computer vision tracking processing with
The tangential great error of nonlinear distortion bring, is demarcated, and carry out duplex feedback adjusting by finding out distortion factor,
Eliminate radial and tangential nonlinear distortion.
In addition, the application also provides a kind of camera shooting machine detecting device applied to safety monitoring slope, which includes:
Object space coordinate system establishes unit, which describes the space coordinate where target, gets target
Body parameter, and it is stored in database；
Slope monitoring image capturing unit acquires the slope monitoring shadow in side slope region to be monitored from least two different angles
Picture；
Identifying body parameter extraction unit is identified the identification of body according to slope monitoring image, extracts identifying body parameter；Mark
Knowledge body is the preset identification model based on color, shapes and sizes parameter；
Slope displacement variable quantity monitoring unit, identifying body parameter and objective body parameter are compared, if do not become
Change, it is lasting to monitor, if there is a change, then three-dimensional computations are carried out, obtains slope displacement variable quantity.
The scheme of existing detection slope stability requires the devices such as embedding strain or sensing inside side slope, is this this Shen
The side slope surface displacement intelligent high-definition video camera detection device based on image recognition please be propose, is carried out using binocular vision technology
Image recognition, it is different from traditional " point " measurement monitoring method, is a kind of non-contact technology based on " face " measurement, has energy
The advantages of reaching rapid survey structure surface three dimension coordinate, and realizing remote network monitoring.
It should be emphasized that term "comprises/comprising" refers to the presence of feature, element, step or component when using herein, but simultaneously
It is not excluded for the presence or additional of one or more other features, element, step or component.
In addition, method of the invention be not limited to specifications described in time sequencing execute, can also according to it
His time sequencing, concurrently or independently execute.Therefore, the execution sequence of method described in this specification is not to this hair
Bright technical scope is construed as limiting.
Although being had been disclosed above by the description to specific embodiments of the present invention to the present invention, it answers
The understanding, above-mentioned all embodiments and example are exemplary, and not restrictive.Those skilled in the art can be in institute
Design is to various modifications of the invention, improvement or equivalent in attached spirit and scope of the claims.These modification, improve or
Person's equivalent should also be as being to be considered as included in protection scope of the present invention.
Claims (9)
1. a kind of video camera detection method applied to safety monitoring slope, this method comprises:
Step 1: establishing object space coordinate system, which describes the space coordinate where target, gets objective body
Parameter, and it is stored in database；
Step 2: a video camera being at least respectively set at left and right sides of side slope, is denoted as left video camera and right video camera；Left camera shooting
Machine and right video camera acquire the slope monitoring image in side slope region to be monitored from least two different angles；
Step 3: image being monitored by left and right cameras side slope and carries out Feature Points Matching, and is identified the identification of body, is extracted
Identifying body parameter；Identifying body is the preset identification module based on color, shapes and sizes parameter；
Step 4: identifying body parameter and objective body parameter being compared, if do not changed, show not to be subjected to displacement；After
Continuous monitoring then carries out three-dimensional computations, obtains slope displacement variable quantity up to changing.
2. video camera detection method according to claim 1, it is characterised in that: carry out characteristic point by left and right cameras
With comprising the following processes:
Process A1: the characteristic point coordinate geometry of left images, respectively L and R are inputted；
Process A2: traversal set L calculates each characteristic point and correspond to polar curve equation in right image, then on calculating R each point to often
The distance of polar curve equation, if distance is less than 1 pixel, then it is assumed that the point is candidate point；
Process A3: if candidate point number is 1, left figure picture point will be changed and be put into set Lt, candidate point is put into set Rt, LtAnd RtFor
The set of characteristic points of successful match；If candidate point number is greater than 1, left figure picture point will be changed and be put into set Lh, candidate point set
It is put into set Rh；
Process A4: traversal set Lh, to each characteristic pointIn LtIn take Euclidean distance nearest characteristic point as left detent pointCorresponding RtIn characteristic point be right detent point
Process A5: left detent point is calculatedWith characteristic pointVector angleAnd Euclidean distance
Process A6: traversal characteristic pointCorresponding RhIn candidate point set, calculate right detent pointWith the vector of each candidate point
AngleAnd Euclidean distanceAnd withWithIt makes the difference；Minimum angles threshold values 10, minimum range threshold values 10, in candidate are set
The candidate point for finding difference minimum and branch threshold requirements of waiting a moment o'clock is match point, and left figure characteristic point is put into Lt, right figure
R is put into pointt；Until having traversed Lh, can be obtained and left figure characteristic point matched right figure characteristic point one by one.
3. video camera detection method according to claim 1, it is characterised in that: in step 3, the identification of identifying body includes such as
Lower process:
Process B1: left and right cameras carries out Feature Points Matching, gets the image containing identifying body, progress color differentiation first:
Get colors bright-coloured identifying body, is different from the environment of surrounding；HIS color segmentation can be used to particular color area in color image
It is effectively extracted in domain；
Process B2: and then carry out shape discrimination: binaryzation, edge detection, contours extract, rectangular detection are carried out to identifying body；
Process B3: then carry out size differentiation, using model matching method calculate similarity, similarity it is high be determined as target pair
As.
4. video camera detection method according to claim 1, it is characterised in that:
In step 4, three-dimensional computations are the three-dimensional coordinates for calculating monitoring point comprising:
Process 1: establishing photo coordinate system O1xy, indicates are as follows:
In formula: ku, kv are respectively length factor of the unit pixel in x-axis, y-axis direction；(Uo, Vo) is photo coordinate system original place
The image coordinate of O1；
Establish camera coordinate system OcXcYcZc；
Establish object space coordinate system: described with translation vector t and spin matrix R object space coordinate system and camera coordinate system it
Between relationship, if homogeneous coordinates of the point P under camera coordinate system in space are (Xc, Yc, Zc) T, the two exists as follows
Relationship:
Wherein t is 3 dimension translation vectors, and R is 3*4 unit matrix, Ot=(0,0,0)T, M is relational matrix, be two coordinate systems it
Between tie；
Process 2: the three-dimensional coordinate of monitoring point calculates:
Relationship between the image coordinate and object space coordinate system of monitoring point is expressed as:
Assuming that M is the parameter matrix of 3*4 rank, may be expressed as:
If two camera parameters matrixes are respectively MA and MB, can obtain:
(ma 11—uama 31)Xw+(ma 12—uama 32)Yw+(ma 13—uama 33) Zw=(uama 34—ma 14)；
(ma 21—vama 31)Xw+(ma 22—vama 32)Yw+(ma 23—vama 33) Zw=(vama 34—ma 24)；
(mb 11—ubma 31)Xw+(mb 12—ubmb 32)Yw+(mb 13—ubmb 33) Zw=(ubmb 34—mb 14)；
(mb 21—vbmb 31)Xw+(mb 22—vbmb 32)Yw+(mb 23—vbmb 33) Zw=(vbmb 34—mb 24)；
Using least square method solving equations, the object space three-dimensional coordinate (x, y, z) of monitoring objective point can be calculated；When finding out
After the three-dimensional coordinate of each observation mark, it is compared with the coordinate under original state respectively, i.e. acquisition side slope surface hair
Raw actual displacement amount.
5. video camera detection method according to claim 1, it is characterised in that: in step 4, three-dimensional computations include following mistake
Journey:
Observation is identified with according to pin hole projection:
Similarly, positioning identifier has:
WhereinProjection array area for observation mark i in t moment, i.e., the area identified in image,Observation mark i exists
The real area of t moment；Indicate that t moment video camera identifies the distance of i to observation；It is positioning identifier 1 in t moment
Projection array area；Distance of the video camera to positioning identifier 1；F is focal length of camera；
Simultaneous has:
Similarly, have for positioning identifier n:
Under projected coordinate system, B1、BnWith AiProjection is away from being respectively as follows:
Enable depth difference:
Raw footage is obtained by correction, then obtains correct depth value by projecting scaling relationship；
A under object space coordinate system is solved with positioning identifier projector distance battle array, depth difference battle array by camera distance battle array, observation markiThree
It ties up coordinate (x, y, z).
6. video camera detection method according to claim 5, it is characterised in that: the calculating process of slope displacement variable quantity is such as
Under:
The starting material space coordinate of monitoring point is stored as (X on side slope surfacen0, Yn0, Zn0), the coordinate of t moment is (Xnt, Ynt,
Znt), then have:
In formula: SnFor the displacement of n-th of measuring point；S is side slope surface displacement；N=1,2 ..., N, wherein N is total measuring point number.
7. video camera detection method according to claim 1, it is characterised in that: before step 2 acquires image, there are also right
The process that video camera is demarcated, the process include:
Observation mark is directly fabricated to scaling board M, can directly determine homography matrix a H, an orthogonal matrix R；H=
H column vector is unfolded, is had by sMR:
H=[h1 h2 h3]=sM [r1 r2t]；
Solution obtains:
Wherein R is orthogonal matrix, r1And r2It is unit orthogonal vectors, i.e., its inner product is 0, mould 1；
Then haveAcquire following formula:
h1M-TM-1h2=0；
If B=M-TM-1, M is the matrix of 3*3, i.e. B is a symmetrical matrix, can be unfolded are as follows:
Wherein B is symmetrical matrix, and only 6 times are original, Matrix Multiplication is opened available:
Available N number of identifying body image simultaneously, stacking these equations has: Vb=0；
Wherein V is the matrix of 2N*6, and as K > 2, b has solution；Camera intrinsic parameter can be directly obtained from the closing solution of B matrix
fx、fy、cx、cyValue；
It has been calculated before outer parameter:
r3=r1*r2；
Acquire the calibration that video camera is completed after the inside and outside parameter of video camera.
8. video camera detection method according to claim 1, it is characterised in that: the identifying body is that vertical angles are identified
Square module.
9. a kind of camera shooting machine detecting device applied to safety monitoring slope, the device include:
Object space coordinate system establishes unit, which describes the space coordinate where target, gets objective body ginseng
Number, and it is stored in database；
Slope monitoring image capturing unit, is at least respectively set a video camera at left and right sides of side slope, be denoted as left video camera and
Right video camera；Left video camera and right video camera acquire the slope monitoring shadow in side slope region to be monitored from least two different angles
Picture；
Identifying body parameter extraction unit carries out Feature Points Matching by left and right cameras, and is identified the identification of body, extracts mark
Know body parameter；Identifying body is the preset identification module based on color, shapes and sizes parameter；Slope displacement variable quantity prison
Unit is surveyed, identifying body parameter and objective body parameter are compared, if do not changed, show not to be subjected to displacement；Continue
Monitoring then carries out three-dimensional computations, obtains slope displacement variable quantity up to changing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910599471.1A CN110307790A (en) | 2019-07-04 | 2019-07-04 | Camera shooting machine detecting device and method applied to safety monitoring slope |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910599471.1A CN110307790A (en) | 2019-07-04 | 2019-07-04 | Camera shooting machine detecting device and method applied to safety monitoring slope |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110307790A true CN110307790A (en) | 2019-10-08 |
Family
ID=68078031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910599471.1A Pending CN110307790A (en) | 2019-07-04 | 2019-07-04 | Camera shooting machine detecting device and method applied to safety monitoring slope |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110307790A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111739093A (en) * | 2020-06-18 | 2020-10-02 | 中国神华能源股份有限公司国华电力分公司 | Method and device for monitoring displacement of steam pipeline and electronic equipment |
| CN111753712A (en) * | 2020-06-22 | 2020-10-09 | 中国电力科学研究院有限公司 | Method, system and equipment for monitoring safety of power production personnel |
| CN112012062A (en) * | 2020-09-01 | 2020-12-01 | 天津动网信息科技有限公司 | Rail and roadbed offset three-coordinate monitoring device and method |
| CN112325789A (en) * | 2021-01-04 | 2021-02-05 | 中南大学 | Method for measuring deformation and displacement in model test based on image processing |
| CN112734833A (en) * | 2020-12-10 | 2021-04-30 | 安徽理工大学 | Matlab-based model target coordinate feedback system |
| CN112819770A (en) * | 2021-01-26 | 2021-05-18 | 中国人民解放军陆军军医大学第一附属医院 | Iodine contrast agent allergy monitoring method and system |
| CN112833838A (en) * | 2020-09-30 | 2021-05-25 | 昆明理工大学 | Polymorphic space three-dimensional displacement monitoring device and displacement calculation method |
| CN112964192A (en) * | 2021-04-07 | 2021-06-15 | 中煤科工集团重庆研究院有限公司 | Engineering measurement online calibration method and system based on image video |
| CN113280750A (en) * | 2021-06-09 | 2021-08-20 | 武汉大学 | Three-dimensional deformation monitoring method and device |
| CN113295095A (en) * | 2021-07-27 | 2021-08-24 | 成都理工大学 | High fill side slope geotechnical centrifugal model measurement control system |
| CN113344955A (en) * | 2021-05-25 | 2021-09-03 | 云南凤云高速公路有限公司 | System and method for accurately detecting slope by fuzzy edge |
| CN113899405A (en) * | 2021-10-20 | 2022-01-07 | 中煤科工集团沈阳设计研究院有限公司 | Integrated online slope intelligent monitoring and early warning system and early warning method |
| CN114250796A (en) * | 2021-12-30 | 2022-03-29 | 北京铁科特种工程技术有限公司 | Composite anchoring part for emergency construction and displacement monitoring and construction method |
| CN114333243A (en) * | 2021-12-21 | 2022-04-12 | 长江三峡勘测研究院有限公司（武汉） | Landslide monitoring and early warning method, device, medium, electronic equipment and terminal |
| CN114663501A (en) * | 2022-05-17 | 2022-06-24 | 深圳市华世智能科技有限公司 | Geological disaster early warning method based on feature code image pose detection |
| CN114719770A (en) * | 2022-04-02 | 2022-07-08 | 基康仪器股份有限公司 | Deformation monitoring method and device based on image recognition and spatial positioning technology |
| CN115063372A (en) * | 2022-06-24 | 2022-09-16 | 深圳特科动力技术有限公司 | Method for monitoring deformation of facility body |
| CN115388769A (en) * | 2022-08-17 | 2022-11-25 | 成都飞机工业（集团）有限责任公司 | Method, device, equipment and medium for monitoring tool state in real time |
| CN115797784A (en) * | 2023-02-06 | 2023-03-14 | 中铁二十五局集团第二工程有限公司 | Road slope geological monitoring method and monitoring system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002055310A (en) * | 2000-08-07 | 2002-02-20 | Fukushima Kikaku Jimusho:Kk | Prism stereoscopic viewing device |
| CN102831751A (en) * | 2012-09-04 | 2012-12-19 | 广东省公路管理局 | Road high-dangerous slope monitoring method based on double-camera imaging technology |
| CN104501735A (en) * | 2014-12-23 | 2015-04-08 | 大连理工大学 | Method for observing three-dimensional deformation of side slope by utilizing circular marking points |
-
2019
- 2019-07-04 CN CN201910599471.1A patent/CN110307790A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002055310A (en) * | 2000-08-07 | 2002-02-20 | Fukushima Kikaku Jimusho:Kk | Prism stereoscopic viewing device |
| CN102831751A (en) * | 2012-09-04 | 2012-12-19 | 广东省公路管理局 | Road high-dangerous slope monitoring method based on double-camera imaging technology |
| CN104501735A (en) * | 2014-12-23 | 2015-04-08 | 大连理工大学 | Method for observing three-dimensional deformation of side slope by utilizing circular marking points |
Non-Patent Citations (4)
| Title |
|---|
| 彭艺: "基于图像识别的边坡表面位移检测方法研究", 《中国优秀硕士学位论文全文数据库（电子期刊）工程科技II辑》 * |
| 李传朋: "基于机器视觉和深度学习的目标识别与抓取定位", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
| 沈慧杰: "基于单目视觉的摄像机定位方法的研究", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
| 许薛军 等: "基于双目视觉的公路边坡表面位移识别方法", 《中国公路学报》 * |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111739093B (en) * | 2020-06-18 | 2024-05-10 | 中国神华能源股份有限公司国华电力分公司 | Steam pipeline displacement monitoring method and device and electronic equipment |
| CN111739093A (en) * | 2020-06-18 | 2020-10-02 | 中国神华能源股份有限公司国华电力分公司 | Method and device for monitoring displacement of steam pipeline and electronic equipment |
| CN111753712A (en) * | 2020-06-22 | 2020-10-09 | 中国电力科学研究院有限公司 | Method, system and equipment for monitoring safety of power production personnel |
| CN112012062A (en) * | 2020-09-01 | 2020-12-01 | 天津动网信息科技有限公司 | Rail and roadbed offset three-coordinate monitoring device and method |
| CN112833838A (en) * | 2020-09-30 | 2021-05-25 | 昆明理工大学 | Polymorphic space three-dimensional displacement monitoring device and displacement calculation method |
| CN112833838B (en) * | 2020-09-30 | 2023-04-28 | 昆明理工大学 | Polymorphic space three-dimensional displacement monitoring device and displacement calculation method |
| CN112734833A (en) * | 2020-12-10 | 2021-04-30 | 安徽理工大学 | Matlab-based model target coordinate feedback system |
| CN112325789A (en) * | 2021-01-04 | 2021-02-05 | 中南大学 | Method for measuring deformation and displacement in model test based on image processing |
| CN112819770A (en) * | 2021-01-26 | 2021-05-18 | 中国人民解放军陆军军医大学第一附属医院 | Iodine contrast agent allergy monitoring method and system |
| CN112819770B (en) * | 2021-01-26 | 2022-11-22 | 中国人民解放军陆军军医大学第一附属医院 | Iodine contrast agent allergy monitoring method and system |
| CN112964192A (en) * | 2021-04-07 | 2021-06-15 | 中煤科工集团重庆研究院有限公司 | Engineering measurement online calibration method and system based on image video |
| CN113344955A (en) * | 2021-05-25 | 2021-09-03 | 云南凤云高速公路有限公司 | System and method for accurately detecting slope by fuzzy edge |
| CN113280750A (en) * | 2021-06-09 | 2021-08-20 | 武汉大学 | Three-dimensional deformation monitoring method and device |
| CN113280750B (en) * | 2021-06-09 | 2022-08-30 | 武汉大学 | Three-dimensional deformation monitoring method and device |
| CN113295095B (en) * | 2021-07-27 | 2021-10-15 | 成都理工大学 | High fill side slope geotechnical centrifugal model measurement control system |
| CN113295095A (en) * | 2021-07-27 | 2021-08-24 | 成都理工大学 | High fill side slope geotechnical centrifugal model measurement control system |
| CN113899405A (en) * | 2021-10-20 | 2022-01-07 | 中煤科工集团沈阳设计研究院有限公司 | Integrated online slope intelligent monitoring and early warning system and early warning method |
| CN114333243A (en) * | 2021-12-21 | 2022-04-12 | 长江三峡勘测研究院有限公司（武汉） | Landslide monitoring and early warning method, device, medium, electronic equipment and terminal |
| CN114250796A (en) * | 2021-12-30 | 2022-03-29 | 北京铁科特种工程技术有限公司 | Composite anchoring part for emergency construction and displacement monitoring and construction method |
| CN114719770A (en) * | 2022-04-02 | 2022-07-08 | 基康仪器股份有限公司 | Deformation monitoring method and device based on image recognition and spatial positioning technology |
| CN114719770B (en) * | 2022-04-02 | 2024-04-02 | 基康仪器股份有限公司 | Deformation monitoring method and device based on image recognition and space positioning technology |
| CN114663501A (en) * | 2022-05-17 | 2022-06-24 | 深圳市华世智能科技有限公司 | Geological disaster early warning method based on feature code image pose detection |
| CN115063372A (en) * | 2022-06-24 | 2022-09-16 | 深圳特科动力技术有限公司 | Method for monitoring deformation of facility body |
| CN115063372B (en) * | 2022-06-24 | 2024-05-31 | 深圳特科动力技术有限公司 | Facility main body deformation monitoring method |
| CN115388769A (en) * | 2022-08-17 | 2022-11-25 | 成都飞机工业（集团）有限责任公司 | Method, device, equipment and medium for monitoring tool state in real time |
| CN115388769B (en) * | 2022-08-17 | 2024-06-07 | 成都飞机工业（集团）有限责任公司 | Real-time tool state monitoring method, device, equipment and medium |
| CN115797784A (en) * | 2023-02-06 | 2023-03-14 | 中铁二十五局集团第二工程有限公司 | Road slope geological monitoring method and monitoring system |
| CN115797784B (en) * | 2023-02-06 | 2023-04-28 | 中铁二十五局集团第二工程有限公司 | Highway slope geological monitoring method and monitoring system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110307790A (en) | Camera shooting machine detecting device and method applied to safety monitoring slope | |
| CN105678742B (en) | A kind of underwater camera scaling method | |
| CN105627926B (en) | Four-camera group planar array feature point three-dimensional measurement system and measurement method | |
| US10019838B2 (en) | Human body three-dimensional imaging method and system | |
| KR100513055B1 (en) | 3D scene model generation apparatus and method through the fusion of disparity map and depth map | |
| CN108010085A (en) | Target identification method based on binocular Visible Light Camera Yu thermal infrared camera | |
| CN102445165B (en) | Stereo vision measurement method based on single-frame color coding grating | |
| CN107907048A (en) | A kind of binocular stereo vision method for three-dimensional measurement based on line-structured light scanning | |
| CN109859272A (en) | A kind of auto-focusing binocular camera scaling method and device | |
| CN104539928B (en) | A kind of grating stereo printing image combining method | |
| CN109579695B (en) | Part measuring method based on heterogeneous stereoscopic vision | |
| JP2004340840A (en) | Distance measuring device, distance measuring method and distance measuring program | |
| CN109443209A (en) | A kind of line-structured light system calibrating method based on homography matrix | |
| CN112945141B (en) | Structured light rapid imaging method and system based on micro-lens array | |
| CN110044374A (en) | A kind of method and odometer of the monocular vision measurement mileage based on characteristics of image | |
| CN106767526A (en) | A kind of colored multi-thread 3-d laser measurement method based on the projection of laser MEMS galvanometers | |
| CN111429527A (en) | Method and system for automatically calibrating external parameters of vehicle-mounted camera | |
| CA3233222A1 (en) | Method, apparatus and device for photogrammetry, and storage medium | |
| Jazayeri et al. | Interest operators for feature‐based matching in close range photogrammetry | |
| CN108279677A (en) | Track machine people's detection method based on binocular vision sensor | |
| CN109373912A (en) | A kind of non-contact six-freedom displacement measurement method based on binocular vision | |
| CN112184793B (en) | Depth data processing method and device and readable storage medium | |
| CN112802114B (en) | Multi-vision sensor fusion device, method thereof and electronic equipment | |
| KR101538014B1 (en) | Converting method photographic images to 3d images through 3d modeling | |
| CN110033491A (en) | A kind of camera calibration 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191008 |