CN109801216A - The quick joining method of Tunnel testing image - Google Patents
The quick joining method of Tunnel testing image Download PDFInfo
- Publication number
- CN109801216A CN109801216A CN201811561093.XA CN201811561093A CN109801216A CN 109801216 A CN109801216 A CN 109801216A CN 201811561093 A CN201811561093 A CN 201811561093A CN 109801216 A CN109801216 A CN 109801216A
- Authority
- CN
- China
- Prior art keywords
- image
- tunnel
- section
- camera
- cross
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to the quick splicings for the tunnel image that Tunnel testing technical field more particularly to Tunnel testing vehicle obtain, and find to use for subsequent tunnel defect.The invention discloses a kind of quick joining method of Tunnel testing image, key step includes the cutting of the calibration of tunnel model image, the unification of the resolution ratio of tunnel image, picture superposition, the even light of image, image mosaic and cross-section image.The present invention can fast implement the splicing of tunnel cross-section image, it is high to splice obtained tunnel cross-section picture contrast, the splicing precision of tunnel image is high, and subsequent the tunnel image that the cross-section image that splicing obtains synthesizes big visual field can be used for automatically extracting for disease.The present invention can greatly save manpower, while facilitate greatly improving for tunnel defect detection efficiency.
Description
Technical field
The present invention relates to the fast Speed Pinyins for the tunnel image that Tunnel testing technical field more particularly to Tunnel testing vehicle obtain
It connects, finds to use for subsequent tunnel defect.
Background technique
At present in infrastructure, such as the Defect inspection of highway pavement diseases detection, road and rail tunnel or bridge.Tradition
Detection means be for example freeway tunnel to be detected based on artificial.By the method for artificial detection, not only to traffic
It impacts, with greater need for a large amount of manpower and time is spent, tends to rely on the experience of testing staff to determine the degree of disease.
Occur replacing the Tunnel testing vehicle of artificial detection at present, Tunnel testing vehicle is equipped with camera, with detection vehicle
Mobile, section obtains tunnel-liner surface image one by one, then splices to a sheet by a sheet tunnel cross-section image, generates big visual field
Tunnel image.In this way, it is possible to detect tunnel using image recognition using tunnel defect automatic identification detection image software
Disease.This method greatly reduces manual intervention, detection efficiency height, is greatly saved relative to traditional artificial detection algorithm
Workload and personnel's investment, greatly improve the efficiency of tunnel defect detection.
The tunnel image of big visual field is made of a sheet by a sheet tunnel cross-section, and obtains tunnel cross-section figure to accelerate to detect vehicle
The speed of picture detects and is typically provided with multiple cameras on vehicle and works at the same time, and each camera obtains the office of different angle on same section
Portion's image.Thus tunnel cross-section is that multiple image mosaics obtained by multiple cameras form.Therefore the tunnel image of big visual field its
To be carried out as a sheet by a sheet tunnel image composed by constantly splicing in fact, the influential effect of splicing the processing such as subsequent Defect inspection
The splicing of process, tunnel image is even more important.
Summary of the invention
Technical problem to be solved by the invention is to provide a kind of quick joining methods of Tunnel testing image, are used for tunnel
The tunnel-liner face image that each CCD area array cameras obtains on road detection vehicle is accurately and fast spliced.
In order to solve the above technical problems, the present invention provides a kind of quick joining method of Tunnel testing image, including following
Step:
(1) tunnel model is demarcated, calibration includes: the offset parameter calculated between tunnel model image;Determination is more
The corresponding relationship of a CCD camera and laser point;Obtain subpoint of the laser scanner in camera support plane to camera away from
From;
(2) Tunnel testing vehicle to actual tunnel carry out data acquisition, the data of acquisition include tunnel-liner face image and
The corresponding laser data of every group of image;The corresponding object distance of true tunnel image at this time is calculated according to laser data, then basis
Object distance and focal length calculate the resolution ratio of every image, and the unification of resolution ratio is carried out to every image;
(3) tunnel image mosaic is handled: practical object distance is calculated according to the overlap distance between the image of peg model
The pixel deviated between lower tunnel image, by one group of tunnel image mosaic at a tunnel cross-section;
(4) the tunnel cross-section image obtained to splicing is cut: first passing through the pixel of camera triggering spacing and image
Resolution ratio calculates the cutting amount of section, is then reduced lap between tunnel cross-section image according to the section cutting amount
Fall, guarantees that subsequent obtained big visual field tunnel image does not have repeating part appearance, obtain the tunnel cross-section figure that a spelling connects
Picture.
Further, the quick joining method of the Tunnel testing image to described the following steps are included: differentiate
Every image degree of the comparing enhancing processing of rate after reunification keeps it visually more clear to improve the contrast of every image
It is clear.
More preferably, the quick joining method of the Tunnel testing image is further comprising the steps of: to every tunnel figure
After image contrast enhancing processing, dodging is carried out, greyscale transformation of the dodging based on overlapping region makes adjacent image
Brightness reach unanimity.
Further, in the tunnel image mosaic processing, the characteristic matching of image overlapping region is introduced between image
Offset pixels constrained, guarantee tunnel image mosaic precision.
The quick joining method of the Tunnel testing image, further comprising the steps of:
Traverse effective tunneling data that Tunnel testing vehicle obtains, these data are successively carried out the resolution ratio it is unified,
Contrast enhancing, the even light of image, image mosaic and image cropping treatment process, until all effective tunnel image is
Splicing is completed.
The invention has the advantages that the splicing of tunnel cross-section image can be fast implemented, the tunnel cross-section spliced
Picture contrast is high, and the splicing precision of tunnel image is high, and the subsequent obtained cross-section image that can will splice synthesize big view
The tunnel image of field being automatically extracted for disease.Manpower can be greatly saved, tunnel defect detection efficiency is improved.
Detailed description of the invention
Technical solution of the present invention is further described in detail with reference to the accompanying drawings and detailed description.
Fig. 1 offset parameter demarcation flow figure between model image.
Fig. 2 is the geometrical relationship figure of laser scanner and camera.
Fig. 3 is the CCD pixel mapping relations figure of pixel resolution and camera.
Fig. 4 is the geometric representation of actual tunnel object distance and tunnel model.
Fig. 5 is the cross-section image with overlaid pixel.
The tunnel Fig. 6 image Fast stitching algorithm flow chart.
Specific embodiment
Present embodiment carries out tunnel using a kind of Tunnel testing multiple-sensor integration platform of the applicant's invention
The acquisition of image is (referring to granted patent " a kind of Tunnel testing multiple-sensor integration platform "-Patent No. that applicant is previous
201510870169.7).Tunnel testing multiple-sensor integration platform is equipped with 16 CCD cameras, a laser scanner.
In conjunction with shown in Fig. 1 to Fig. 6, the Fast stitching algorithm of tunnel image carries out the calibration to tunnel model, then first
It acquires actual tunnel-liner face data and resolution ratio unification is carried out according to laser data, image degree of comparing is improved later
And dodging with guarantee that tunnel image detail is clear and image between brightness it is consistent, then according to actual tunnel lining
Object distance by the image Overlapping parameters under tunnel model calculate under the object distance to determine the offset picture between actual tunnel image
Prime number mesh, tunnel image mosaic be section after section is cut to guarantee there is no repeat region appearance between section.It connects down
The embodiment of each section to be specifically described.
It before carrying out tunnel-liner face image mosaic, needs to demarcate tunnel model, to obtain 16 under model
Offset parameter between image acquired in a CCD camera is used to calculate when subsequent splicing practical to tunnel-liner face image real
Offset parameter between border tunnel image.Calibration is divided into the calculating of offset parameter, camera and laser point between tunnel model image
The subpoint that corresponding relationship is determining and laser scanner is in camera support plane is to three portions such as the acquisitions of distance of camera
Point.
Calculating for offset parameter between tunnel model image, first according to the model analysis document pair of Tunnel testing vehicle
Tunnel model image carries out the unification of resolution ratio, and 16 tunnel model images are then imported Photoshop, chooses between image
Same place, the offset pixels of image are obtained according to same place, the weight between image can easily be obtained by Photoshop
Fold-over element, by the overlaid pixel between image by formula
Li=overlapi×Basic_resolution
It is scaled actual length, wherein LiCorresponding physical length, overlap between imageiWeight between image
Fold-over element, Basic_resolution are benchmark resolution ratio, and process is as shown in Figure 1.
Then corresponding relationship between laser and image is obtained, is blocked corresponding camera using baffle, by the laser number of acquisition
According to distance is scaled, laser distance data are then shown by application program matlab, because baffle is an apparent plane,
Therefore the part of a plane is shown as on matlab is exactly laser point number corresponding to the plane, by 16 CCD camera models
After laser point number in enclosing determines, point number is exported in txt respectively by camera serial number.
The last one process of calibration is laser scanner between the subpoint and each camera in camera support plane
The measurement of distance is the geometrical relationship between laser scanner and camera as shown in Figure 2, and laser scanner is in camera support plane
On subpoint be O, if one of camera is A, laser scanner B, need using laser to obtain object distance SA ' to tunnel
The distance BS in road face subtracts A ' B, that is, subtracts AO, is known according to laser data BS, therefore only requires to obtain AO, utilizes hook
Stock theorem, using total station survey AB and OB, so that it may and then obtain AO, 16 cameras are carried out seeking it respectively to sweep to laser
Retouch the distance of instrument subpoint in camera support plane.So far, between the tunnel model image of calibration process offset parameter meter
Calculation, camera are with the determination of laser point correspondence and laser scanner at a distance from the subpoint to camera in camera support plane
Determination completed.
After demarcating to tunnel model image, start the acquisition for carrying out actual tunnel-liner face image data, to obtaining
The actual tunnel-liner face taken obtains the distance of laser scanner to tunnel face according to its laser data.Laser scanner
Data are range data, and the laser scanner on Tunnel testing vehicle acquires the range data of 541 points altogether, and each point is according to order
Number consecutively can from 541 points because the corresponding relationship of image and laser point has been determined during calibration
Laser readily to extract the corresponding laser point data of every image, for a tunnel image, in image range
Point data is it has been determined that calculate the average value of its laser point data, i.e., the average value of the distance of these point representatives is as the image
Place tunnel face is at a distance from laser scanner, which is the BS in Fig. 2, here for the abnormal data in image range
It will do it identification, in practical applications, it is found that its value of the data of laser point exception is 0, therefore for being worth the laser point data for being 0
It is not included in averaged.The process of calibration has calculated laser scanner in camera support plane
A ' B is known to subpoint in i.e. Fig. 2 at a distance from camera, then by formula di=BSi-A′BiIt can be concluded that each tunnel figure
As corresponding object distance di, then differentiated according to the pixel that the size of the focal length of camera lens and CCD pixel calculates each image
Rate, as shown in figure 3, f is focal length, d is object distance, and CCD is the size of camera CCD pixel, then the pixel resolution of image
By
The corresponding pixel resolution resolution of 16 tunnel images can be found outi.Then a benchmark is selected
Image progress resampling in tunnel is made its pixel resolution be equal to reference resolution to complete the system of image resolution ratio by resolution ratio
One.
Followed by tunnel image degree of the comparing enhancing of resolution ratio after reunification is handled, equilibrium is taken the photograph using segmentation is imitative here
Carry out the contrast enhancement processing of image.It is that dodging is carried out to each image later, brightness between adjacent image is made to tend to one
It causes.Here even light is carried out in a manner of image between any two even light.Here for using left image as standard video, by right image
Carrying out processing makes its brightness reach unanimity with left image.Using one kind based on the matched gray scale affine transformation of partitioned histogram to the right side
The piecewise affine that piece carries out by stages is balanced, so that reaching makes right and the consistent effect of standard video brightness, and by equal
Weighing apparatusization, right details are enhanced.
In order to guarantee that atural object gray scale of the same race is consistent to the greatest extent after even light, left images overlapping region is counted respectively first
Gray scale accumulative histogram because the two overlay region have identical atural object, it is same on gray accumulation histogram
The atural object of the corresponding gray scale of probability should be it is identical, can choose specific two probability positions respectively from left and right two images
On select gray scale L1, L2 and R1, R2, L1 and L2 are left gray scales, and R1 and R2 are right gray scales, and L1 and R1 is of the same name
Gray scale, L2 and R2 are gray scales of the same name.Right is carried out based on the corresponding greyscale transformation of gray scale of the same name, right gray value after transformation
Range should and R1 identical as left transformed to L1, R2 is transformed to L2.To through based on the corresponding progress greyscale transformation of gray scale of the same name
The right progress by stages statistics normalization accumulative histogram of processing, the subinterval of the right image after greyscale transformation should be with a left side
Image is identical, thus can guarantee that the endpoint in each section is atural object gray scale of the same name, carries out in this way to each section corresponding straight
Side's figure matching is equivalent to the atural object being between atural object gray scale of the same name to gray scale and is matched, that is, assumes right warp based on of the same name
Its gray scale interval is by [b after point gray scale stretchingk, bk+1] it is converted into [ak, ak+1], to right gray area after left and gray scale stretching
Between [ak, ak+1] accumulation histogram statistics is normalized, horizontal axis is gray value, and it is total that the longitudinal axis is that each gray level accounts for gray scale interval
The probability of pixel number, it is seen that the gray scale stretching processing to right also ensures the correspondence of horizontal axis tonal range.If left gray scale
Section [ak, ak+1] normalization accumulation histogram distribution are as follows:
S=T (r)
Gray scale stretching treated right gray scale interval [ak, ak+1] normalization accumulation histogram distribution are as follows:
V=G (z)
Wherein r, z are former ash degree, and right gray scale z can be indicated are as follows:
Z=G-1(v)
Pass through section [ak, ak+1] Histogram Matching, we can look for it is equal with v to a s, then:
Z=G-1(T(r))
Therefore for right gray scale z in left same gray scale interval [ak, ak+1] have a gray scale r corresponding, pass through
Determine the r transformation section [x locating in left piecewise affine equilibrium treatmentk, xk+1], so that it is determined that right of drawn processing
Gray scale is the transformation slope m when pixel of z carries out affine equilibriumk, affine transformation is carried out, finally realizes right after stretch processing
Section [ak, ak+1] affine equilibrium, successively the gray scale of rest interval is proceeded as described above until completing the imitative of whole picture image
Penetrate equilibrium.Piecewise affine equilibrium for right is in fact one and finds left corresponding grey scale by by stages Histogram Matching,
And then the gray scale according to the affine transformation relationship of left corresponding grey scale to right carries out affine equilibrium treatment, by straight based on subregion
Scheme matched gray scale affine transformation and avoid to two width caused by right after the gray scale stretching directly equilibrium of progress piecewise affine in side
The excessive difference of the gray scale of atural object of the same race in image.So far the dodging between adjacent image is completed.For other images two
Even light and the above process between two are similarly.
Following step is exactly to carry out splicing to the image by enhancing and dodging, and when splicing needs to calculate
Actual tunnel image offset pixels horizontal and vertical between any two.According to actual object distance object corresponding with tunnel model image
Away from the geometrical relationship between, viewing field of camera angle, the calibrated horizontal and vertical offset distance of tunnel model image is converted to
Under the object distance of actual tunnel image, and offset pixels are calculated according to the relationship with pixel resolution.Actual tunnel object distance with
The geometric representation of tunnel model such as Fig. 4, because camera is closer between any two on actual tunnel detection vehicle, it is believed that camera is thrown
Shadow is parallel in the plane approximation in tunnel face.As shown in figure 4, θAAnd θBThe respectively half of the field angle of camera A and camera B,
fAAnd fBThe respectively focal length of camera A and camera B, D1And D2When respectively camera B corresponding calibration the object distance of tunnel model image with
And the object distance of the actual tunnel image of camera B acquisition, the overlap length of actual tunnel image is approximately: known to geometrical relationship
Overlap=(D2-D1)(tanθA+tanθB)+L
It then can will be weighed according to image resolution ratio in the hope of the overlap length of tunnel image between any two according to above-mentioned relation
Folded length is scaled pixel number, and then obtains the pixel number that image is overlapped between any two.According to the superposition image prime number acquired
List when mesh can calculate splicing between image answers the translation parameters in gust Homography, then introduces the overlapping between image
The characteristic matching in region is modified the translation parameters in Homography according to accurate matched result, thus further
The precision spliced between image is improved, elimination is carried out using Laplacian Pyramid method when splicing two-by-two between image and is connect
Seam processing is to realize that the gray scale of image intersection seamlessly transits.Tunnel image is spliced to obtain tunnel cross-section image.?
In actual application, it can choose the need for not introducing the application that matched mode is spliced with satisfaction under lower accuracy requirement
It wants.
Final step is that the tunnel cross-section image obtained to splicing is cut, the triggering of the CCD camera of Tunnel testing vehicle
Spacing is 460mm, i.e., will expose once every 460mm, obtains one group 16 and opens tunnel image, and 460mm is differentiated according to pixel
Rate conversion is the number of pixels of image, has relationship as shown in Figure 5 with the height of tunnel cross-section.Rows is the height of cross-section image, r1
Indicate the corresponding pixel number of triggering spacing 460mm, r2 indicates the overlaid pixel between two sections, it was found from its relationship:
R2=Rows-r1
The matching being equally also introduced between section here constrains it, to guarantee the accuracy of cutting amount r2, in reality
It can according to need the acquisition for choosing whether that cutting amount is carried out using the matched mode between section in the application of border.
Traverse effective tunneling data that Tunnel testing vehicle obtains, these data are successively carried out above-mentioned resolution ratio it is unified,
Contrast enhancing, the even light of image, image mosaic and image cropping treatment process, until all effective tunnel image is
Splicing is completed.
It should be noted last that the above specific embodiment is only used to illustrate the technical scheme of the present invention and not to limit it,
Although being described the invention in detail referring to preferred embodiment, those skilled in the art should understand that, it can be right
Technical solution of the present invention is modified or replaced equivalently, without departing from the spirit and scope of the technical solution of the present invention,
It is intended to be within the scope of the claims of the invention.
Claims (5)
1. a kind of quick joining method of Tunnel testing image, which comprises the following steps:
(1) tunnel model is demarcated, calibration includes: the offset parameter calculated between tunnel model image;Determine multiple CCD
The corresponding relationship of camera and laser point;Obtain the distance of subpoint of the laser scanner in camera support plane to camera;
(2) Tunnel testing vehicle carries out data acquisition to actual tunnel, and the data of acquisition include tunnel-liner face image and every group
The corresponding laser data of image;The corresponding object distance of true tunnel image at this time is calculated according to laser data, then according to object distance
And focal length, the resolution ratio of every image is calculated, the unification of resolution ratio is carried out to every image;
(3) tunnel image mosaic is handled: tunnel under practical object distance is calculated according to the overlap distance between the image of peg model
The pixel deviated between road image, by one group of tunnel image mosaic at a tunnel cross-section;
(4) the tunnel cross-section image obtained to splicing is cut: the pixel for first passing through camera triggering spacing and image is differentiated
Rate calculates the cutting amount of section, then reduces by lap between tunnel cross-section image according to the section cutting amount, obtain
The tunnel cross-section image connected to a spelling.
2. the quick joining method of Tunnel testing image according to claim 1, which comprises the following steps:
To every image degree of the comparing enhancing processing of the progress resolution ratio after reunification.
3. the quick joining method of Tunnel testing image according to claim 2, which comprises the following steps:
After the processing of every tunnel picture superposition, dodging, ash of the dodging based on overlapping region are carried out
Degree transformation, makes the brightness of adjacent image reach unanimity.
4. the quick joining method of the Tunnel testing image according to claim 3, which is characterized in that the tunnel figure
As in splicing, the characteristic matching for introducing image overlapping region constrains the offset pixels between image.
5. the quick joining method of Tunnel testing image according to claim 4, which is characterized in that further include following step
It is rapid:
Effective tunneling data that Tunnel testing vehicle obtains is traversed, these data are successively carried out with the resolution ratio unification, comparison
Spend enhancing, the even light of image, image mosaic and image cropping treatment process, spliced up to all effective tunnel images
It completes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811561093.XA CN109801216B (en) | 2018-12-20 | 2018-12-20 | Rapid splicing method for tunnel detection images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811561093.XA CN109801216B (en) | 2018-12-20 | 2018-12-20 | Rapid splicing method for tunnel detection images |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109801216A true CN109801216A (en) | 2019-05-24 |
CN109801216B CN109801216B (en) | 2023-01-03 |
Family
ID=66557194
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811561093.XA Active CN109801216B (en) | 2018-12-20 | 2018-12-20 | Rapid splicing method for tunnel detection images |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109801216B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110378856A (en) * | 2019-07-22 | 2019-10-25 | 福建农林大学 | A kind of tunnel surface two-dimensional laser image enhancement processing method |
CN110764106A (en) * | 2019-10-09 | 2020-02-07 | 中交一公局集团有限公司 | Construction method for assisting shield interval slope and line adjustment measurement by adopting laser radar |
CN111583108A (en) * | 2020-04-20 | 2020-08-25 | 北京新桥技术发展有限公司 | Tunnel lining surface linear array image TOF fusion splicing method and device and storage medium |
CN111707668A (en) * | 2020-05-28 | 2020-09-25 | 武汉武大卓越科技有限责任公司 | Tunnel detection and image processing method based on sequence image |
CN112435170A (en) * | 2020-12-04 | 2021-03-02 | 安徽圭目机器人有限公司 | Tunnel vault image splicing method |
CN113310987A (en) * | 2020-02-26 | 2021-08-27 | 保定市天河电子技术有限公司 | Tunnel lining surface detection system and method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH04353199A (en) * | 1991-05-30 | 1992-12-08 | Hitachi Constr Mach Co Ltd | Attitude detection method and automatic built-up device of segment |
JP2015049765A (en) * | 2013-09-03 | 2015-03-16 | 公益財団法人鉄道総合技術研究所 | Method of correcting distortion of tunnel lining surface image |
CN105550995A (en) * | 2016-01-27 | 2016-05-04 | 武汉武大卓越科技有限责任公司 | Tunnel image splicing method and system |
CN108267096A (en) * | 2017-12-25 | 2018-07-10 | 中铁科学技术开发公司 | Railway tunnel lining surface disease rapid detection system |
-
2018
- 2018-12-20 CN CN201811561093.XA patent/CN109801216B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH04353199A (en) * | 1991-05-30 | 1992-12-08 | Hitachi Constr Mach Co Ltd | Attitude detection method and automatic built-up device of segment |
JP2015049765A (en) * | 2013-09-03 | 2015-03-16 | 公益財団法人鉄道総合技術研究所 | Method of correcting distortion of tunnel lining surface image |
CN105550995A (en) * | 2016-01-27 | 2016-05-04 | 武汉武大卓越科技有限责任公司 | Tunnel image splicing method and system |
CN108267096A (en) * | 2017-12-25 | 2018-07-10 | 中铁科学技术开发公司 | Railway tunnel lining surface disease rapid detection system |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110378856A (en) * | 2019-07-22 | 2019-10-25 | 福建农林大学 | A kind of tunnel surface two-dimensional laser image enhancement processing method |
CN110764106A (en) * | 2019-10-09 | 2020-02-07 | 中交一公局集团有限公司 | Construction method for assisting shield interval slope and line adjustment measurement by adopting laser radar |
CN113310987A (en) * | 2020-02-26 | 2021-08-27 | 保定市天河电子技术有限公司 | Tunnel lining surface detection system and method |
CN111583108A (en) * | 2020-04-20 | 2020-08-25 | 北京新桥技术发展有限公司 | Tunnel lining surface linear array image TOF fusion splicing method and device and storage medium |
CN111583108B (en) * | 2020-04-20 | 2020-12-18 | 北京新桥技术发展有限公司 | Tunnel lining surface linear array image TOF fusion splicing method and device and storage medium |
CN111707668A (en) * | 2020-05-28 | 2020-09-25 | 武汉武大卓越科技有限责任公司 | Tunnel detection and image processing method based on sequence image |
CN111707668B (en) * | 2020-05-28 | 2023-11-17 | 武汉光谷卓越科技股份有限公司 | Tunnel detection and image processing method based on sequence images |
CN112435170A (en) * | 2020-12-04 | 2021-03-02 | 安徽圭目机器人有限公司 | Tunnel vault image splicing method |
CN112435170B (en) * | 2020-12-04 | 2023-11-03 | 安徽圭目机器人有限公司 | Tunnel vault image splicing method |
Also Published As
Publication number | Publication date |
---|---|
CN109801216B (en) | 2023-01-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109801216A (en) | The quick joining method of Tunnel testing image | |
CN111260615B (en) | Laser and machine vision fusion-based method for detecting apparent diseases of unmanned aerial vehicle bridge | |
CN103902953B (en) | A kind of screen detecting system and method | |
CN111583411A (en) | Three-dimensional model building method based on oblique photography | |
CN113608378B (en) | Full-automatic defect detection method and system based on LCD (liquid crystal display) process | |
CN106996748A (en) | A kind of wheel footpath measuring method based on binocular vision | |
JPWO2020059706A1 (en) | Learning data collection device, learning data collection method, and program | |
CN105787870A (en) | Graphic image splicing fusion system | |
Ioli et al. | UAV photogrammetry for metric evaluation of concrete bridge cracks | |
CN115330684A (en) | Underwater structure apparent defect detection method based on binocular vision and line structured light | |
CN113375555A (en) | Power line clamp measuring method and system based on mobile phone image | |
CN111862082A (en) | Train brake pad thickness rechecking method and system thereof | |
JPH05215547A (en) | Method for determining corresponding points between stereo images | |
JP2020160626A (en) | Construction drawing creation support system | |
US20220076428A1 (en) | Product positioning method | |
CA2230197C (en) | Strand orientation sensing | |
Farhang et al. | Rapid estimation method for vertical deflection measurement of building components based on image processing techniques | |
JP2961140B2 (en) | Image processing method | |
JPH10318732A (en) | Shape measuring device and image formation apparatus of shape measurement | |
CN107860933B (en) | Digital image-based automatic detection method and device for fiber content in textile | |
JPH10318715A (en) | Image measuring apparatus | |
JPH10320560A (en) | Image measuring instrument | |
CN118015068B (en) | Pavement structure depth prediction method and device, terminal equipment and medium | |
JPH0981778A (en) | Device and method for modeling | |
CN117197135B (en) | Wall surface flatness detection method and system based on laser point cloud |
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 | ||
CB02 | Change of applicant information |
Address after: 430223 Hubei science and Technology Park, East Lake Development Zone, Wuhan, China Applicant after: Wuhan Optical Valley excellence Technology Co.,Ltd. Address before: 430223 No.6, 4th Road, Wuda Science Park, Donghu New Technology Development Zone, Wuhan City, Hubei Province Applicant before: WUHAN WUDA ZOYON SCIENCE AND TECHNOLOGY Co.,Ltd. |
|
CB02 | Change of applicant information | ||
GR01 | Patent grant | ||
GR01 | Patent grant |