CN108109112A - A kind of tunnel spread figure splicing parameter processing method based on Sift features - Google Patents
A kind of tunnel spread figure splicing parameter processing method based on Sift features Download PDFInfo
- Publication number
- CN108109112A CN108109112A CN201810038940.8A CN201810038940A CN108109112A CN 108109112 A CN108109112 A CN 108109112A CN 201810038940 A CN201810038940 A CN 201810038940A CN 108109112 A CN108109112 A CN 108109112A
- Authority
- CN
- China
- Prior art keywords
- denoted
- image
- idx
- point
- matching
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 18
- 238000000034 method Methods 0.000 claims abstract description 11
- 230000001594 aberrant effect Effects 0.000 claims abstract description 4
- 101100243951 Caenorhabditis elegans pie-1 gene Proteins 0.000 claims description 14
- 238000000205 computational method Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010187 selection method Methods 0.000 claims description 3
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- 230000001737 promoting effect Effects 0.000 claims 1
- 238000012360 testing method Methods 0.000 abstract description 4
- 238000013519 translation Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 6
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011017 operating method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30132—Masonry; Concrete
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30244—Camera pose
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention splices parameter processing method for a kind of tunnel spread figure based on Sift features, and for splicing to image captured by Tunnel testing vehicle, the method comprises the following steps：S1, selection image are effectively matched scope；S2, acquisition currently carry out matched image；S3, brightness of image is promoted according to the gray average of image；S4, the Sift characteristic points for obtaining image simultaneously reject pseudo- match point；S5, Mismatching point is further rejected using the translation specifications of image characteristic point；S6, the number for calculating the average of distance, variance and match point between matching characteristic point after correcting, if meeting decision threshold, it is believed that current two figure successful match otherwise since step 2, reselect matching figure, until meeting condition position；S7, the degree of overlapping for calculating two images, that is, splice parameter；S8, based on experience value corrects aberrant splicing parameter.The present invention can accurately and rapidly obtain tunnel spread figure splicing parameter.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of tunnel spread figure splicing based on Sift features
Parameter processing method.
Background technology
Tunnel Engineering is often the node engineering of highway, railway, track traffic and urban road, tunnel it is quick
Development brings huge facility and economic interests.It after tunnel is built up, is influenced, will necessarily be caused each by each side such as natural causes
Kind disease or damage, therefore the maintenance in tunnel is increasingly paid attention to.
Traditional tunnel defect detection is mostly detected dependent on visual inspection or artificial instrument, but by overhaul efficiency
Restriction, conventional method can not be achieved substantially comprehensively covering exhaustively detect, can not also ensure the periodical, timely of result
Property and objectivity.At present, main method carries out Image Acquisition using picture pick-up device to body surface, by means of image procossing skill
Art is handled and identified to image, and disease geo-radar image, the type for marking disease and position can be shown in tunnel spread figure
Etc. features.It is required to meet the precision of images, is repeatedly shot frequently with multiple cameras, then again splice image,
Finally obtain tunnel spread figure.
Currently used two types of merging algorithm for images classification：（1）Based on the relevant stitching algorithm in region, rely on
In the gray value of image to be spliced, this method is often because brightness of image, contrast, provincial characteristics are closer to and cause
Splicing identification.（2）The relevant stitching algorithm of feature based.By the use of characteristics of image as standard, to the correspondence of image lap
Characteristic area scans for matching, and stability is preferable.But existing algorithm is usually present the situation of error hiding, in tunnel
The image that feature is single, similarity is high, effect is worse, this will seriously affect the effect of image mosaic.
Therefore, it is necessary to the image shot using Tunnel testing vehicle, circumferential same by multiple cameras with degree of overlapping
Step shooting, longitudinally through equidistant this feature of triggering camera, the method for improving Feature Points Matching obtains the spelling of tunnel spread figure
Connect parameter.
The content of the invention
It is main to solve the object of the present invention is to provide a kind of tunnel spread figure splicing parameter processing method based on Sift features
The technical issues of high region recognition of similarity is poor in certainly current image mosaic.The thinking of the present invention is to multiple in Tunnel testing vehicle
The equidistant triggering of camera, the tunnel-liner image of synchronization sync pulse jamming are analyzed, and the progress of splicing is divided in detail
Analysis and research, it is proposed that effectively solve method, can quickly and accurately calculate the splicing parameter of tunnel spread figure.
The present invention can be achieved through the following technical solutions：A kind of tunnel spread figure splicing parameter based on Sift features
Processing method comprises the following steps：
S1, selection image are effectively matched scope：Start image number is denoted as N1, terminates picture number and is denoted as N2, wherein N1<
N2；Under normal conditions, the matching range of selection includes 50 images, can be maximum so in the case of it there is matching image
It is possible to obtain matching image, in the case of there is no matching image, it can jump out in time, save the time to the full extent.
S2, acquisition currently carry out matched image, and the concrete operations of the step S2 are as follows：
（1）Camera numbers are set, Base Serial Number is denoted as C1, and finish number is denoted as C2, wherein C1<C2；
（2）Selection currently carries out matched image, and selection method is as follows：
（a）For longitudinal spliced parameter, it is necessary to be chosen in the image of same camera shooting, camera numbers are denoted as Cam_
Idx chooses first figure as source figure, and number is denoted as Pic1_Idx, and image is denoted as Pic1_Src, and in addition a picture number is
Pic2_Idx, image are denoted as Pic2_Src.Wherein, Pic2_Idx=Pic1_Idx+1, Pic1_Idx[N1, N2-1], Cam_
Idx[C1, C2]；
（b）For circumferential splicing parameter, it is necessary to be chosen in the image of adjacent cameras shooting, first figure is chosen as source
Figure, camera numbers Cam1_Idx, picture number Pic_Idx, image are denoted as Pic1_Src, in addition the camera of an image
Number is Cam2_Idx, and picture number is similarly Pic_Idx, and image is denoted as Pic2_Src, wherein Cam2_Idx=Cam1_Idx-
1, Pic_Idx, Cam1_Idx。
S3, brightness of image is promoted according to the gray average of image, the concrete operations of the step S3 are as follows：
（1）The gradation of image average of artwork is calculated, original image is set to I, and average is denoted as MeanValue；
（2）If the desired value of gradation of image average is TargetValue, luminance factor is denoted as coef, and computational methods are as follows：
(1)
(2)
Wherein, TargetValue is the luminance mean value of image after adjustment brightness, and the scope being generally worth is located between 120 ~ 150,
The value of TargetValue is too small, and brightness is inadequate, also poor to the effect of characteristics of image enhancing, and the value of TargetValue is too big, figure
As excessively bright, it will cause missing image part details, therefore it is preferable to be set to currency effect；
（3）Image after adjustment brightness is denoted as I_adlight,
(3)
So, the image selected in step S2 is directed to, is denoted as Pic1 and Pic2, the width and high score of image after brightness adjustment respectively
W, H are not denoted as it（Pic1 is consistent with the size of Pic2）.
S4, the Sift characteristic points for obtaining image simultaneously reject pseudo- match point, and the operation of the step S4 is as follows：
（1）Using sift characteristics algorithms, obtain Pic1 and Pic2 feature point coordinates be denoted as respectively KeyPoint1,
KeyPoint2, description are denoted as Descriptors1, Descriptors2 respectively；
（2）Characteristic matching is carried out using flann algorithms, the characteristic point after matching is denoted as Matches1；
（3）Mismatching point is deleted using RANSAC algorithms, matching characteristic point is denoted as Matches2.
S5, Mismatching point is further rejected using the translation specifications of image characteristic point, the operation of the step S5 is as follows：
(1) all matching characteristic points pair are traveled through, note source characteristic point is P1, and coordinate points information is（P1X, P1Y）, target feature point
For P2, coordinate points information is（P2X, P2Y）；
(2) according to X-direction and the distance feature of Y-direction, Mismatching point is rejected, is as follows：
(a) for longitudinal spliced：The distance of its Y-direction is calculated, is denoted as Dis_Y, Dis_Y=| | P1Y-P2Y | |；Given threshold is
Threshold_Y, if Dis_Y<Threshold_Y, then it is assumed that otherwise this picks match point match point successful match by this
It removes, the value of threshold_Y is set to 0.2*H;
(b) splice for circumferential direction：The distance of its X-direction is calculated, is denoted as Dis_X, Dis_X=| | P1X-P2X | |；Given threshold is
Threshold_X, if Dis_X<Threshold_X, then it is assumed that otherwise this picks match point match point successful match by this
It removes, the value of threshold_X is set to 0.2*W;
(3) characteristic point of successful match is recorded again, the X-coordinate at image Pic1 midpoints is denoted as Point1X, and Y-coordinate is denoted as
The X-coordinate at Point1Y, image Pic2 midpoint is denoted as Point2X, and Y-coordinate is denoted as Point2Y.
S6, the number for calculating the average of distance, variance and match point between matching characteristic point after correcting, judge if meeting
Threshold value, it is believed that current two figure successful match otherwise since step S2, reselect matching figure, are until meeting condition
Only, the concrete operations of the step S6 are as follows：
（1）The number of matching characteristic point pair is obtained as N, number threshold value is set as Tn, if N<Tn, then it is assumed that matching is unsuccessful, from
S2 steps reselect matching image；Here the setting of Tn, is generally set to 6, if the value of Tn is excessive, condition will be excessively harsh,
It is less to meet the matching image of condition, really matching image will be missed；It, will be wrong there are some if the value of Tn is smaller
The characteristic point matched somebody with somebody, but still can meet screening conditions, therefore, the step is very crucial；
（2）The distance of match point X-direction and Y-direction in calculation procedure 5
(a) for longitudinal spliced, X-direction distance is denoted as VerDisX, and Y-direction distance is denoted as VerDisY, and computational methods are as follows：
VerDisX(i)=（4）
VerDisY(i)= （5）
Wherein,；
(b) splice for circumferential direction, X-direction distance is denoted as CirDisX, and Y-direction distance is denoted as CirDisY, and computational methods are as follows：
CirDisX(i)=（6）
CirDisY(i)= （7）
Wherein,；
（3）The average and variance of X-direction distance and Y-direction distance are calculated, is denoted as meanX, varX, meanY, varY respectively；
（4）It is right（3）Middle parameter setting threshold value is respectively TMX, TVX, TMY, TVY, when meeting the following conditions
（8）
When, then it is assumed that present image successful match otherwise since step S2, reselects matching figure, until meeting condition.
S7, the degree of overlapping for calculating two images, that is, splice parameter, image degree of overlapping is denoted as overlap_degree：（a）It is right
Splice in circumferential direction, overlap_degree=meanX/W；（b）For longitudinal spliced, overlap_degree=meanY/H.
S8, based on experience value corrects aberrant splicing parameter, and circumferential direction splicing parameter is denoted as Circle_overlap, and longitudinal direction is spelled
It connects parameter and is denoted as Vertical_overlap, when characteristic matching meets condition, record the splicing parameter being currently calculated;If
All images have been traveled through, have not still met the image of condition, then overlap_degree=0.Before calculating parameter, obtain
One group of empirical value is taken, as overlap_degree=0, is substituted with empirical value.
The beneficial effects of the invention are as follows：
（1）The present invention is matched using image characteristic point, while using matching judgement is carried out the characteristics of captured image, is not had
There are complicated image processing operations, calculate precise and high efficiency；
（2）The present invention is modeled according to the physical arrangement in tunnel and the installation site of camera, full-automatic and be easily achieved, surely
Qualitative and accuracy rate higher；
（3）The present invention handles anomaly parameter, it can be ensured that splices the accuracy of parameter.
Description of the drawings
Fig. 1 is the operating procedure flow diagram of the method for the present invention.
Fig. 2 is the matched step schematic diagram of feature of present invention.
Specific embodiment
The invention will be further described in the following with reference to the drawings and specific embodiments.
With reference to the accompanying drawings 1, the present invention is a kind of tunnel spread figure splicing parameter processing method based on Sift features, described
Tunnel spread figure splicing parameter immediate processing method comprises the following steps S1~S8：
S1, selection image are effectively matched scope：Start image number is denoted as S1, terminates picture number and is denoted as N2, wherein N1<
N2；Under normal conditions, the matching range of selection includes 50 images, can be maximum so in the case of it there is matching image
It is possible to obtain matching image, in the case of there is no matching image, it can jump out in time, save the time to the full extent；
The range parameter that is effectively matched of image can be defined as follows：
Struct VivadeIdxRange
{
int Start_Idx;// be effectively matched scope start image number
int End_Idx;// it is effectively matched the end picture number of scope
}。
S2, acquisition currently carry out matched image, and the concrete operations of the step S2 are as follows：
（1）Camera numbers are set, Base Serial Number is denoted as C1, and finish number is denoted as C2, wherein C1<C2；
（2）Selection currently carries out matched image, and selection method is as follows：
（a）For longitudinal spliced parameter, it is necessary to be chosen in the image of same camera shooting, camera numbers are denoted as Cam_
Idx chooses first figure as source figure, and number is denoted as Pic1_Idx, and image is denoted as Pic1_Src, and in addition a picture number is
Pic2_Idx, image are denoted as Pic2_Src.Wherein, Pic2_Idx=Pic1_Idx+1, Pic1_Idx[N1, N2-1], Cam_Idx[C1, C2]；
（b）For circumferential splicing parameter, it is necessary to be chosen in the image of adjacent cameras shooting, first figure is chosen as source
Figure, camera numbers Cam1_Idx, picture number Pic_Idx, image are denoted as Pic1_Src, in addition the camera of an image
Number is Cam2_Idx, and picture number is similarly Pic_Idx, and image is denoted as Pic2_Src, wherein Cam2_Idx=Cam1_Idx-
1, Pic_Idx, Cam1_Idx；
The parameter of current matching image can be defined as follows：
Struct MatchImgPara
{
int Cam1_Idx;The camera numbers of // source figure
int Cam2_Idx;The camera numbers of // target figure
int Pic1_Idx;The picture number of // source figure
int Pic2_Idx;The picture number of // target figure
}。
S3, brightness of image is promoted according to the gray average of image, the concrete operations of the step S3 are as follows：
（1）The gradation of image average of artwork is calculated, original image is set to I, and average is denoted as MeanValue；
（2）If the desired value of gradation of image average is TargetValue, luminance factor is denoted as coef, and computational methods are as follows：
Wherein, TargetValue is the luminance mean value of image after adjustment brightness, and the scope being generally worth is located between 120 ~ 150,
The value of TargetValue is too small, and brightness is inadequate, also poor to the effect of characteristics of image enhancing, and the value of TargetValue is too big, figure
As excessively bright, it will cause missing image part details, therefore it is preferable to be set to currency effect；
（3）Image after adjustment brightness is denoted as I_adlight,
So, the image selected in step S2 is directed to, is denoted as Pic1 and Pic2, the width and high score of image after brightness adjustment respectively
W, H are not denoted as it（Pic1 is consistent with the size of Pic2）.
S4, the Sift characteristic points for obtaining image simultaneously reject pseudo- match point, and the operation of the step S4 is as follows：
（1）Using sift characteristics algorithms, obtain Pic1 and Pic2 feature point coordinates be denoted as respectively KeyPoint1,
KeyPoint2, description are denoted as Descriptors1, Descriptors2 respectively；
（2）Characteristic matching is carried out using flann algorithms, the characteristic point after matching is denoted as Matches1；
（3）Mismatching point is deleted using RANSAC algorithms, matching characteristic point is denoted as Matches2.
S5, Mismatching point is further rejected using the translation specifications of image characteristic point, the operation of the step S5 is as follows：
(1) all matching characteristic points pair are traveled through, note source characteristic point is P1, and coordinate points information is（P1X, P1Y）, target feature point
For P2, coordinate points information is（P2X, P2Y）；
(2) according to X-direction and the distance feature of Y-direction, Mismatching point is rejected, is as follows：
(a) for longitudinal spliced：The distance of its Y-direction is calculated, is denoted as Dis_Y, Dis_Y=| | P1Y-P2Y | |.Given threshold is
Threshold_Y, if Dis_Y<Threshold_Y, then it is assumed that otherwise this picks match point match point successful match by this
It removes, the value of threshold_Y is set to 0.2*H;
(b) splice for circumferential direction：The distance of its X-direction is calculated, is denoted as Dis_X, Dis_X=| | P1X-P2X | |.Given threshold is
Threshold_X, if Dis_X<Threshold_X, then it is assumed that otherwise this picks match point match point successful match by this
It removes, the value of threshold_X is set to 0.2*W;
(3) characteristic point of successful match is recorded again, the X-coordinate at image Pic1 midpoints is denoted as Point1X, and Y-coordinate is denoted as
The X-coordinate at Point1Y, image Pic2 midpoint is denoted as Point2X, and Y-coordinate is denoted as Point2Y.
S6, the number for calculating the average of distance, variance and match point between matching characteristic point after correcting, judge if meeting
Threshold value, it is believed that current two figure successful match otherwise since step S2, reselect matching figure, are until meeting condition
Only, the concrete operations of the step S6 are as follows：
（1）The number of matching characteristic point pair is obtained as N, number threshold value is set as Tn, if N<Tn, then it is assumed that matching is unsuccessful, from
S2 steps reselect matching image；Here the setting of Tn, is generally set to 6, if the value of Tn is excessive, condition will be excessively harsh,
It is less to meet the matching image of condition, really matching image will be missed；It, will be wrong there are some if the value of Tn is smaller
The characteristic point matched somebody with somebody, but still can meet screening conditions, therefore, the step is very crucial；
（2）The distance of match point X-direction and Y-direction in calculation procedure 5
(a) for longitudinal spliced, X-direction distance is denoted as VerDisX, and Y-direction distance is denoted as VerDisY, and computational methods are as follows：
VerDisX(i)=
VerDisY(i)=
Wherein,；
(b) splice for circumferential direction, X-direction distance is denoted as CirDisX, and Y-direction distance is denoted as CirDisY, and computational methods are as follows：
CirDisX(i)=
CirDisY(i)=
Wherein,；
（3）The average and variance of X-direction distance and Y-direction distance are calculated, is denoted as meanX, varX, meanY, varY respectively；
（4）It is right（3）Middle parameter setting threshold value is respectively TMX, TVX, TMY, TVY, when meeting the following conditions
When, then it is assumed that present image successful match otherwise since step S2, reselects matching figure, until meeting condition；
The parameter of current matching image distance can be defined as follows：
Struct DisMeanVar
{
double meanX;The average of // image X-direction
double meanY;The average of // image Y-direction
double varX;The variance of // image X-direction
double varY;The variance of // image Y-direction
}。
S7, the degree of overlapping for calculating two images, that is, splice parameter, image degree of overlapping is denoted as overlap_degree：（a）It is right
Splice in circumferential direction, overlap_degree=meanX/W；（b）For longitudinal spliced, overlap_degree=meanY/H.
S8, based on experience value corrects aberrant splicing parameter, and circumferential direction splicing parameter is denoted as Circle_overlap, and longitudinal direction is spelled
It connects parameter and is denoted as Vertical_overlap, when characteristic matching meets condition, record the splicing parameter being currently calculated;If
All images have been traveled through, have not still met the image of condition, then overlap_degree=0.Before calculating parameter, obtain
One group of empirical value is taken, as overlap_degree=0, is substituted with empirical value.
Since tunnel spread figure is spliced along the direct of travel of Tunnel testing vehicle, camera position in shooting process
Fixed and camera is set a distance triggering, so spread figure in tunnel of the present invention splicing parameter immediate processing method is basis
Longitudinal spliced image realizes in circumferential dislocation and circumferential stitching image longitudinally fluctuating smaller principle, and according to each phase
The longitudinal spliced parameter of machine is consistent and the circumferential splicing parameter of adjacent cameras unanimously determines the splicing parameter of tunnel spread figure.
It is above the preferred embodiment of the present invention.It should be appreciated that those of ordinary skill in the art are without creative
Work, which according to the present invention can conceive, makes many modifications and variations.Therefore, all technician in the art are according to this
The design of invention passes through the available technical side of logical analysis, reasoning, or a limited experiment on the basis of existing technology
Case, all should be in as the protection domain required by claims of the present invention.
Claims (8)
- A kind of 1. tunnel spread figure splicing parameter processing method based on Sift features, which is characterized in that the tunnel spread Figure splicing parameter immediate processing method comprises the following steps：S1, the scope that is effectively matched for selecting image, start image number are denoted as N1, terminate picture number and be denoted as N2, wherein N1< N2；S2, acquisition currently carry out matched image, and source seal is Pic1_Src, and target seal is Pic2_Src；S3, brightness of image is promoted according to the gray average of image, Pic1_Src and Pic2_Src are corresponded respectively to after promoting brightness Pic1 and Pic2, the width and height of image are denoted as W, H respectively, and wherein Pic1 is consistent with the size of Pic2；S4, the feature point coordinates of acquisition Pic1 and Pic2 are denoted as KeyPoint1, KeyPoint2 respectively, and description is denoted as respectively Descriptors1, Descriptors2, after rejecting pseudo- match point, matching characteristic point information is denoted as Matches2；S5, Mismatching point is further rejected to the matching characteristic point obtained in step S4, utilizes true match point X-direction and Y side Upward feature rejects Mismatching point, remembers in the figure of source that X-coordinate and Y-coordinate are respectively Point1X, Point1Y, X is sat in target figure Mark and Y-coordinate are respectively Point2X, Point2Y；S6, revised matching characteristic point in step S5 is directed to, further calculates average, the variance of X-direction and Y-direction distance And the number of characteristic point, meet decision condition, then it is assumed that current two figure successful match, otherwise since S2 steps, again Selection matching figure, until meeting condition；S7, the splicing parameter for calculating two images, i.e. degree of overlapping, are denoted as overlap_degree；S8, based on experience value, corrects aberrant splicing parameter, and circumferential direction splicing parameter is denoted as Circle_overlap, longitudinal spliced ginseng Number scale is Vertical_overlap.
- 2. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S2 are as follows：（1）Camera numbers are set, Base Serial Number is denoted as C1, and finish number is denoted as C2, wherein C1<C2;（2）Selection currently carries out matched image, and selection method is as follows：（a）For longitudinal spliced parameter, it is necessary to be chosen in the image of same camera shooting, camera numbers are denoted as Cam_ Idx chooses first figure as source figure, and number is denoted as Pic1_Idx, and image is denoted as Pic1_Src, in addition a picture number note For Pic2_Idx, image is denoted as Pic2_Src.Wherein, Pic2_Idx=Pic1_Idx+1, Pic1_Idx[N1, N2-1], Cam_ Idx[C1, C2]；（b）For circumferential splicing parameter, it is necessary to be chosen in the image of adjacent cameras shooting, first figure is chosen as source Figure, camera numbers Cam1_Idx, picture number Pic_Idx, image are denoted as Pic1_Src, in addition the camera of an image Number is Cam2_Idx, and picture number is similarly Pic_Idx, and image is denoted as Pic2_Src, wherein Cam2_Idx=Cam1_Idx- 1, Pic_Idx, Cam1_Idx[C1+1,C2]。
- 3. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S3 are as follows：（1）The gradation of image average of artwork is calculated, original to be set to I, average is denoted as MeanValue；（2）If the desired value of gradation of image average is TargetValue, luminance factor is denoted as coef, and computational methods are as follows：（3）Image after adjustment brightness is denoted as I_adlight,So, the image selected in step S2 is directed to, Pic1 and Pic2 are denoted as respectively after brightness adjustment.
- 4. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S4 are as follows：（1）Using sift characteristics algorithms, obtain Pic1 and Pic2 feature point coordinates be denoted as respectively KeyPoint1, KeyPoint2, description are denoted as Descriptors1, Descriptors2 respectively；（2）Characteristic matching is carried out using flann algorithms, the characteristic point after matching is denoted as Matches1；（3）Mismatching point is deleted using RANSAC algorithms, matching characteristic point is denoted as Matches2.
- 5. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S5 are as follows：(1) all matching characteristic points pair are traveled through, note source characteristic point is P1, and coordinate points information is（P1X, P1Y）, target feature point For P2, coordinate points information is (P2X, P2Y)；(2) according to X-direction and the distance feature of Y-direction, Mismatching point is rejected, is as follows：(a) for longitudinal spliced：The distance of its Y-direction is calculated, is denoted as Dis_Y, Dis_Y=| | P1Y-P2Y | |;Given threshold is Threshold_Y, if Dis_Y<Threshold_Y, then it is assumed that otherwise this picks match point match point successful match by this It removes；(b) splice for circumferential direction：The distance of its X-direction is calculated, is denoted as Dis_X, Dis_X=| | P1X-P2X | |;Given threshold For threshold_X, if Dis_X<Threshold_X, then it is assumed that otherwise this picks match point match point successful match by this It removes；(3) characteristic point of successful match is recorded again, the X-coordinate at image Pic1 midpoints is denoted as Point1X, and Y-coordinate is denoted as The X-coordinate at Point1Y, image Pic2 midpoint is denoted as Point2X, and Y-coordinate is denoted as Point2Y.
- 6. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S6 are as follows：（1）The number of matching characteristic point pair is obtained as N, number threshold value is set as Tn, if N<Tn, then it is assumed that matching is unsuccessful, from Step S2 reselects matching image；（2）The distance of match point X-direction and Y-direction, specific method are as follows in calculation procedure S5：（a）For longitudinal spliced, X-direction distance is denoted as VerDisX, and Y-direction distance is denoted as VerDisY, and computational methods are as follows：Wherein,；(b) splice for circumferential direction, X-direction distance is denoted as CirDisX, and Y-direction distance is denoted as CirDisY, and computational methods are as follows：Wherein,；（3）Calculate the average and variance of X-direction distance and Y-direction distance, then for longitudinal spliced, calculate VerDisX and The average and variance of VerDisY is spliced for circumferential direction, is calculated the average and variance of CirDisX and CirDisY, is denoted as respectively meanX、varX、meanY、varY；（4）It is right（3）Middle parameter setting threshold value is respectively TMX, TVX, TMY, TVY, when meeting the following conditionsWhen, then it is assumed that present image successful match otherwise since S2, reselects matching figure, until meeting condition.
- 7. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S7 are as follows：Image degree of overlapping is denoted as overlap_degree：（a）Splice for circumferential direction, overlap_degree=meanX/W；（b）It is right In longitudinal spliced, overlap_degree=meanY/H.
- 8. a kind of tunnel spread figure splicing parameter processing method based on Sift features according to claim 1, feature It is, the concrete operations of the step S8 are as follows：When characteristic matching meets condition, the splicing parameter being currently calculated is recorded;If having traveled through all images, still do not have Meet the image of condition, then overlap_degree=0；Before calculating parameter, one group of empirical value is obtained, works as overlap_ During degree=0, substituted with empirical value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810038940.8A CN108109112B (en) | 2018-01-16 | 2018-01-16 | Tunnel layout graph splicing parameter processing method based on Sift characteristic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810038940.8A CN108109112B (en) | 2018-01-16 | 2018-01-16 | Tunnel layout graph splicing parameter processing method based on Sift characteristic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108109112A true CN108109112A (en) | 2018-06-01 |
| CN108109112B CN108109112B (en) | 2021-07-20 |
Family
ID=62219389
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810038940.8A Active CN108109112B (en) | 2018-01-16 | 2018-01-16 | Tunnel layout graph splicing parameter processing method based on Sift characteristic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108109112B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109146791A (en) * | 2018-09-04 | 2019-01-04 | 上海同岩土木工程科技股份有限公司 | A kind of tunnel spread drawing generating method based on area array CCD imaging |
| CN110473236A (en) * | 2019-06-25 | 2019-11-19 | 上海圭目机器人有限公司 | A kind of measurement method of the offset position of road face image detection camera |
| CN117745537A (en) * | 2024-02-21 | 2024-03-22 | 微牌科技(浙江)有限公司 | Tunnel equipment temperature detection method, device, computer equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102088569A (en) * | 2010-10-13 | 2011-06-08 | 首都师范大学 | Sequence image splicing method and system of low-altitude unmanned vehicle |
| CN104021559A (en) * | 2014-06-17 | 2014-09-03 | 西安电子科技大学 | Image registration method based on mutual information and Harris corner point detection |
| US20150294490A1 (en) * | 2014-04-13 | 2015-10-15 | International Business Machines Corporation | System and method for relating corresponding points in images with different viewing angles |
| 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 |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
| CN107093166A (en) * | 2017-04-01 | 2017-08-25 | 华东师范大学 | The seamless joint method of low coincidence factor micro-image |
-
2018
- 2018-01-16 CN CN201810038940.8A patent/CN108109112B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102088569A (en) * | 2010-10-13 | 2011-06-08 | 首都师范大学 | Sequence image splicing method and system of low-altitude unmanned vehicle |
| US20150294490A1 (en) * | 2014-04-13 | 2015-10-15 | International Business Machines Corporation | System and method for relating corresponding points in images with different viewing angles |
| CN104021559A (en) * | 2014-06-17 | 2014-09-03 | 西安电子科技大学 | Image registration method based on mutual information and Harris corner point detection |
| 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 |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
| CN107093166A (en) * | 2017-04-01 | 2017-08-25 | 华东师范大学 | The seamless joint method of low coincidence factor micro-image |
Non-Patent Citations (4)
| Title |
|---|
| MAOSEN WANG 等: "A novel panoramic image stitching algorithm based on ORB", 《PROCEEDINGS OF THE 2017 IEEE INTERNATIONAL CONFERENCE ON APPLIED SYSTEM INNOVATION》 * |
| MING LI 等: "A study on automatic UAV image mosaic method for paroxysmal disaster", 《ISPRS - INTERNATIONAL ARCHIVES OF THE PHOTOGRAMMETRY REMOTE SENSING AND SPATIAL INFORMATION SCIENCES》 * |
| 张静 等: "基于SIFT特征和误匹配逐次去除的图像拼接", 《半导体光电》 * |
| 彭勃宇 等: "面向增强现实的SUSAN-SURF快速匹配算法", 《计算机应用研究》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109146791A (en) * | 2018-09-04 | 2019-01-04 | 上海同岩土木工程科技股份有限公司 | A kind of tunnel spread drawing generating method based on area array CCD imaging |
| CN110473236A (en) * | 2019-06-25 | 2019-11-19 | 上海圭目机器人有限公司 | A kind of measurement method of the offset position of road face image detection camera |
| CN110473236B (en) * | 2019-06-25 | 2022-03-15 | 上海圭目机器人有限公司 | Method for measuring offset position of camera for road surface image detection |
| CN117745537A (en) * | 2024-02-21 | 2024-03-22 | 微牌科技(浙江)有限公司 | Tunnel equipment temperature detection method, device, computer equipment and storage medium |
| CN117745537B (en) * | 2024-02-21 | 2024-05-17 | 微牌科技(浙江)有限公司 | Tunnel equipment temperature detection method, device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108109112B (en) | 2021-07-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107076677B (en) | Inspection apparatus and inspection method | |
| JP5175528B2 (en) | Tunnel lining crack inspection system | |
| CN110892255B (en) | Image processing apparatus, method and storage medium for detecting defect from image | |
| CN108109112A (en) | A kind of tunnel spread figure splicing parameter processing method based on Sift features | |
| CN107345921B (en) | A kind of tire belt fitting quality determining method and system | |
| WO2020110667A1 (en) | Surface defect detecting method, surface defect detecting device, method for manufacturing steel material, steel material quality control method, steel material manufacturing equipment, method for creating surface defect determination model, and surface defect determination model | |
| CN110390256B (en) | Asphalt pavement crack extraction method | |
| CN106501272A (en) | Machine vision scolding tin position detecting system | |
| JP4954469B2 (en) | Appearance inspection method | |
| JP2006170922A (en) | Visual inspection method and its apparatus | |
| CN105354816A (en) | Electronic element positioning method and apparatus | |
| JP3589293B2 (en) | Road white line detection method | |
| JP6908445B2 (en) | Maintenance management method for change detectors and railway equipment parts | |
| JP2011058939A (en) | Apparatus and method for visual inspection | |
| US9305235B1 (en) | System and method for identifying and locating instances of a shape under large variations in linear degrees of freedom and/or stroke widths | |
| JP7130356B2 (en) | Image processing device and maintenance management method for railway equipment parts | |
| CN106645174A (en) | Automatic online visual apparent-defect inspection system for general purpose engine | |
| US10062155B2 (en) | Apparatus and method for detecting defect of image having periodic pattern | |
| CN105469414A (en) | Contour connection method and apparatus | |
| JP3159063B2 (en) | Surface defect inspection equipment | |
| US20080107329A1 (en) | Method of detecting defects of patterns on a semiconductor substrate and apparatus for performing the same | |
| CN109146916A (en) | A kind of moving body track method and device | |
| JP2002358595A (en) | Instrument and method for measuring road traffic stream | |
| JP2010243209A (en) | Defect inspection method and defect detection device | |
| JP2006163662A (en) | Device and method for recognizing number of fingers |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |