CN102930525B - Line matching method based on affine invariant feature and homography - Google Patents
Line matching method based on affine invariant feature and homography Download PDFInfo
- Publication number
- CN102930525B CN102930525B CN201210342566.3A CN201210342566A CN102930525B CN 102930525 B CN102930525 B CN 102930525B CN 201210342566 A CN201210342566 A CN 201210342566A CN 102930525 B CN102930525 B CN 102930525B
- Authority
- CN
- China
- Prior art keywords
- line segment
- same name
- doubtful
- main leaf
- line
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 title claims abstract description 29
- 239000011159 matrix material Substances 0.000 claims description 55
- 230000008878 coupling Effects 0.000 claims description 31
- 238000010168 coupling process Methods 0.000 claims description 31
- 238000005859 coupling reaction Methods 0.000 claims description 31
- 238000000605 extraction Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 239000000284 extract Substances 0.000 claims description 4
- 230000013011 mating Effects 0.000 claims description 3
- 230000007547 defect Effects 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 230000002441 reversible effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- VMXUWOKSQNHOCA-UKTHLTGXSA-N ranitidine Chemical compound [O-][N+](=O)\C=C(/NC)NCCSCC1=CC=C(CN(C)C)O1 VMXUWOKSQNHOCA-UKTHLTGXSA-N 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Abstract
The invention relates to a line matching method based on affine invariant feature and homography. As line matching is lack of an effective geometric constraint of an epipolar line in point matching. A homography constraint is introduced as the geometric constraint of line segment matching to make up the defect that the line segment matching is lack of a strong geometric constraint. The invention additionally discloses a line segment automatic matching method based on the homography constraint. Line segments are transmitted and sleeved among images through the homography constraint, so that the searching difficulty for the line segments of the same name is reduced, and the matching accuracy is improved; the line segments of the same name are backwards searched after primary matching, so that matching errors are removed, and the matching accuracy is further improved. The method achieves the line segment automatic matching for remote-sensing image pairs.
Description
Technical field
The present invention relates to a kind of matching process of remote sensing image, particularly relate to a kind of lines matching method based on affine invariants and homography matrix.
Background technology
Adopt line features as Matching unit, in some application-specific, there is obvious advantage, as in buildings three-dimensional reconstruction (Habib 1998 [1]).This is because: on image, buildings comprises a large amount of straight-line segment; Straight-line segment is easily detected and has clear and definite physical significance more; Line features corresponding point feature has more how describable geometrical constraint, more reliably.
But corresponding point mates, and line match is technically more difficult, main cause is (Schmid and Zisserman 1997 [2]; Baillard and Zisserman 2000 [3]): (1) extracts line segment from image, and line segment ruptures mostly, and the topological relation between line segment is lost; (2) Point matching has the restriction relation that core line geometry is very strong like this, and lines matching lacks such restriction relation.
Current international digital Photogrammetric System is carry out same place search by the constraint of core line geometry mostly, but the constraint of core line can not provide the one-to-one relationship of unique point.Have researchist to propose to utilize corresponding image points to form triangulation network constraint homonymous line hunting zone, the method needs triangle networking in advance, and efficiency is not high.
In multiple view geometry, homography matrix represents the reversible homogeneous transformation between two planes, has been widely used in the fields such as vision measurement, camera calibration, three-dimensional reconstruction, image mosaic, and has played extremely important role wherein.At computer vision field, homography matrix can only be applied to the features convey between two planes in theory, but in Photogrammetry and Remote Sensing field, for aviation image or satellite image, due to the fluctuating of landform or the discrepancy in elevation of atural object very little relative to flying height, homography matrix is also applicable (Schmid and Zisserman 1997 [2]).
List of references:
[1]Habib,A.Motion parameter esmation by tracking stat ionarythree-dimensional straight lines in image sequences.ISPRSJournal of Photogrammetry&Remote Sensing,1998,53,174-182.
[2]Schmid,C.；Zisserman,A.Automatic line matching acrossviews.IEEE Computer Society Conference on Computer Vision andPattern Recognit ion,1997.DOI:10.1109/CVPR.1997.609397.
[3]Baillard C,Zisserman A.A plane-sweep strategy for the 3Dreconstruction of buildings from mult iple images.
International Archives of Photogrammetry and Remote Sensing,2000,33(B2；PART 2),56-62.
Summary of the invention
The technical matters that the present invention solves is: propose a kind of lines matching method based on affine invariants and homography matrix that can improve line match accuracy rate, realize the line segment Auto-matching that remote sensing image is right.
In order to solve the problems of the technologies described above, the technical scheme that the present invention proposes is: a kind of lines matching method based on affine invariants and homography matrix, comprises the following steps:
Step 1, obtain optimum homography matrix between remote sensing image pair---based on the RANSAC iterative algorithm of affine invariants coupling, obtain the optimum homography matrix between remote sensing image pair;
Step 2, transmit remote sensing image between line segment---the line segment extracting remote sensing image centering respectively obtains line chart, and the optimum homography matrix obtained with step 1 is constraint, with one of them line chart for main leaf, another line chart is from sheet, to be delivered on main leaf from the line segment on sheet, original line segment wherein on main leaf is main leaf line segment, and the line segment being delivered to main leaf from sheet is from sheet line segment;
Step 3, determine the sets of line segments doubtful of the same name that main leaf line segment is corresponding---travel through all main leaf line segments, according to the direction between master and slave line segment, Distance geometry degree of overlapping, find all doubtful line segment of the same name corresponding with main leaf line segment from sheet line segment, form the sets of line segments doubtful of the same name of this main leaf line segment;
Step 4, determine from sets of line segments doubtful of the same name corresponding to sheet line segment---travel through all from sheet line segment, according to the direction between master and slave line segment, Distance geometry degree of overlapping, find in main leaf line segment with from all doubtful line segment of the same name corresponding to sheet line segment, form this sets of line segments doubtful of the same name from sheet line segment;
Step 5, rejecting error hiding---compare from the corresponding relation between sheet line segment and its doubtful line segment of the same name in the corresponding relation between the main leaf line segment that step 3 is obtained and its doubtful line segment of the same name and step 4, when main leaf line segment L with step 3, step 4, all there is corresponding relation from sheet line segment L ＇, then described main leaf line segment L with the match is successful from sheet line segment L ＇, otherwise rejects main leaf line segment L and the matching relationship from sheet line segment L ＇.
Innovative point of the present invention is: for the problem of the effective geometrical constraint lacked in lines matching as Point matching center line geometry, introduce the geometrical constraint of homography matrix constraint as line match, to make up in line match the situation lacking strong geometrical constraint, and propose a kind of line segment automatic matching method based on homography matrix constraint, transmission and the fit of line segment between image is achieved by the constraint of homography matrix, reduce line search difficulty of the same name, improve matching accuracy rate; And after preliminary matches completes, reverse search line segment of the same name, thus can error hiding be rejected, further increase matching accuracy rate.
The present invention is that line match is carried out in constraint with homography matrix, and homography matrix is the key of subsequent treatment so accurately.Homography matrix accurately to be obtained between a pair remote sensing image, need to guarantee there is no erroneous matching in the matching process, and retain correct coupling as much as possible.ED-MSER algorithm (the Cheng et al. (2008) utilizing inventor herein to propose, robust affineinvariant feature extraction for image matching, IEEEGeoscience and Remote Sensing Letters, 5 (2)), affine invariants extraction is carried out to remote sensing image, and then characteristic matching, good, the overlapping large image of, texture little for visual angle change can obtain goodish effect, is matched to power very high.But, when face viewpoint angles change greatly, little, the texture of degree of overlapping difficult image time, be matched to power still not ideal enough, some error hiding still exist.Therefore between remote sensing image, the estimation of homography matrix is the key of middle conductor of the present invention coupling, the present invention proposes a kind of RANSAC iterative algorithm based on affine invariants coupling in step 1, achieve homography matrix accurate, reliably obtain, concrete grammar is as follows:
I, use MSER to extract operator, SIFT as feature interpretation operator as affine invariants, carry out the feature extraction of two width images, generating feature vector;
II, do distance operation between the proper vector of two subpictures, the SIFT feature vector obtaining coupling is right;
III, RANSAC method is used to the SIFT feature vector of coupling to processing, the input parameter of RANSAC method is distance threshold, the geometric model of input is the homography matrix of unknown parameters, obtain after process intra-office SIFT feature vector to and the design parameter of homography matrix, when performing this step first RANSAC method distance threshold span for (0,1];
IV, according to the homography matrix that estimates to remote sensing image to mating, calculate matching accuracy rate, described matching accuracy rate is correct number of pairs and the ratio mated sum;
V, progressively increase the distance threshold of RANSAC method, and repeat the 3rd) step to the 4th step until matching accuracy rate declines from 100%, matching accuracy rate be 100% homography matrix corresponding to maximal distance threshold be exactly optimum homography matrix.
Wherein, in step II, during coupling SIFT feature point, when minimal characteristic vector distance is greater than 0.6 with the ratio of time minimal characteristic vector distance, proper vector is apart from the SIFT feature point pair of that minimum a pair SIFT feature point as coupling.
In step IV step, if the degree of overlapping of the right characteristic area of the same name of remote sensing image is greater than 50%, then think both one_to_one corresponding, namely this coupling is correct coupling.
The core concept of the above-mentioned RANSAC iterative algorithm based on affine invariants coupling is: carry out on the basis of affine invariants extraction at ED-MSER, RANSAC algorithm optimization affine invariants is utilized to mate, according to the feature of remote sensing image, geometrical constrain model using homography matrix as RANSAC algorithm, for affine invariants extractive technique obtain the situation that feature is generally territory, face, utilize homography matrix by the feature of the same name of the spurious matches after RANSAC Robust Estimation correct (inliers) to carrying out fit, calculate the degree of overlapping of characteristic area of the same name and upper coupling accuracy calculating feature group in this basis again, to mate the quantitative assessing index of accuracy for iterative processing effect, automatic adjustment distance threshold, instruct the iterative process of RANSAC optimization process, obtain optimum homography matrix.
The line segment extracted remote sensing image usually can in fracture shape, in order to improve line match accuracy, reduce the interference of line match, between completing steps 2 transmits remote sensing image pair after line segment, pre-service can be carried out to the line segment of these fractures.Therefore present invention also offers a kind of to the pretreated method of remote sensing image line segment, specific as follows:
When two line segment angles be less than 4 °, vertical range is not more than 5 pixels, and two line segments have lap: when line segment length difference is more than 50%, retain longer line segment, remove shorter line segment; Otherwise do line segment to merge, get the equal straight line of distance two line segments, and using two line segments farthest two-end-point as the line segment end points after merging.
The present invention provides a kind of method generating main leaf line segment candidate line-segment sets of the same name in step 3, when a certain meet following three conditions from sheet line segment simultaneously time, this be the line segment doubtful of the same name of corresponding main leaf line segment from sheet line segment:
A, with main leaf line segment angle be less than 5 °;
B, with main leaf line segment overlap length account for more than 20% of main leaf line segment length;
C, be not more than 10 pixels to the vertical range of main leaf line segment.
Present invention also offers the method that the doubtful of the same name sets of line segments corresponding to main leaf line segment is optimized, comprise and utilize the similarity of line segment lap gray scale and utilize line segment left and right grey scale signal ratio.
Wherein, the similarity of line segment lap gray scale is utilized to reduce the concrete grammar of doubtful sets of line segments scope of the same name as follows:
Extend out along vertical main leaf line segment direction is symmetrical, formed and extend out rectangle, described in calculating, extend out the remote sensing image average gray M in territory, rectangular foot-print; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the average gray M ＇ that the doubtful line segment of the same name of every bar extends out the remote sensing image in territory, rectangular foot-print in the same way; When M/M ＇ be greater than 2 or be less than 1/2 time, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
Wherein, it is as follows than the concrete grammar reducing doubtful sets of line segments scope of the same name to utilize line segment left and right grey scale signal:
Extend out along vertical main leaf line segment direction is symmetrical, form two rectangles, according to the gray scale of the remote sensing image in territory, two rectangular foot-print, record the side that gray-scale value is high; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the remote sensing image left and right gray scale in the territory, two rectangular foot-print of the doubtful line segment of the same name of every bar in the same way, if the side that doubtful line segment gray-scale value of the same name is a high side position high from main leaf line segment gray-scale value is different, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
The sets of line segments doubtful of the same name that the present invention can also utilize Kmeans clustering algorithm corresponding to main leaf line segment is in step 3 optimized further, and concrete grammar is:
Calculate the vertical range of main leaf line segment to its doubtful line segment of the same name, Kmeans clustering algorithm is utilized to be divided into two classes to doubtful line segment of the same name according to described vertical range, the line segment doubtful of the same name that vertical range is larger is classified as a class automatically, the line segment doubtful of the same name that vertical range value is less is classified as another kind of automatically, rejects line segments doubtful of the same name all in the large class of vertical range.
It should be noted that: the method for the sets of line segments doubtful of the same name that the optimization main leaf line segment provided in step 3 is corresponding, all can correspondingly be applied in step 4, be used for optimizing from sets of line segments doubtful of the same name corresponding to sheet line segment.Doubtful sets of line segments of the same name is optimized, can calculated amount be reduced, improve accuracy rate.
The invention has the beneficial effects as follows: (1) the present invention is directed in lines matching the problem of the effective geometrical constraint lacked as Point matching center line geometry, introduce homography matrix constraint as the geometrical constraint of line match, compensate in line match the situation lacking strong geometrical constraint;
(2) the present invention carries the line segment automatic matching method used based on homography matrix constraint, transmission and the fit of line segment between image is achieved by the constraint of homography matrix, the accurate transmission of line segment across image can be realized, reduce line search difficulty of the same name, improve matching accuracy rate;
(3) the present invention is after preliminary matches completes, and by reverse search line segment of the same name, can reject error hiding, thus further increase matching accuracy rate.
Accompanying drawing explanation
Below in conjunction with accompanying drawing, the lines matching method based on affine invariants and homography matrix of the present invention is described further.
Fig. 1 is one of remote sensing image of the embodiment of the present invention.
Fig. 2 is the remote sensing image two of the embodiment of the present invention.
Fig. 3 is the line chart that Fig. 1 extracts.
Fig. 4 is the line chart that Fig. 2 extracts.
Fig. 5 is the result figure that the line segment of Fig. 4 is delivered to Fig. 3.
Fig. 6 is the line match result figure of Fig. 3.
Fig. 7 is the line match result figure of Fig. 4.
Embodiment
Embodiment
As depicted in figs. 1 and 2, spatial resolution is higher for the experimental data of the present embodiment, and coverage is relatively large, covers the types of ground objects such as building, road, vegetation, bare area.
The lines matching method based on affine invariants and homography matrix of the present embodiment, comprises the following steps:
Step 1, obtain optimum homography matrix between remote sensing image pair.
The present embodiment uses the RANSAC iterative algorithm based on affine invariants coupling to obtain optimum homography matrix between remote sensing image pair, and specific algorithm is as follows:
I, use MSER to extract operator, SIFT as feature interpretation operator as affine invariants, carry out the feature extraction of two width images, generating feature vector;
II, do distance operation between the proper vector of two subpictures, the SIFT feature vector obtaining coupling is right;
During coupling SIFT feature point, when minimal characteristic vector distance is greater than 0.6 with the ratio of time minimal characteristic vector distance, proper vector is apart from the SIFT feature point pair of that minimum a pair SIFT feature point as coupling;
III, RANSAC method is used to the SIFT feature vector of coupling to processing, the input parameter of RANSAC method is distance threshold, the geometric model of input is the homography matrix of unknown parameters, obtain after process intra-office SIFT feature vector to and the design parameter of homography matrix, when performing this step first, the distance threshold of RANSAC method is 1E-6;
IV, according to the homography matrix that estimates to remote sensing image to mating, if the degree of overlapping of the right characteristic area of the same name of remote sensing image is greater than 50%, then think both one_to_one corresponding, namely this coupling is correct coupling; Calculate matching accuracy rate, wherein matching accuracy rate is correct number of pairs and the ratio mated sum;
V, according to formula
(in formula, n is iterations) calculates the value of distance threshold, progressively increase the distance threshold of RANSAC method, and repeat step III to step IV until matching accuracy rate declines from 100%, matching accuracy rate be 100% homography matrix corresponding to maximal distance threshold be exactly optimum homography matrix.
Step 2, transmit remote sensing image between line segment.
The present embodiment utilizes EDISION operator, carries out line segments extraction respectively to Fig. 1 and Fig. 2, and the result obtained respectively as shown in Figure 3 and Figure 4.The optimum homography matrix obtained with step 1 is constraint, and with the image in Fig. 3 for main leaf, in Fig. 4, image is from sheet, will be delivered on the line chart of main leaf from the line segment on sheet, and as shown in Figure 5, the effect of fit is satisfactory; Original line segment wherein on main leaf is main leaf line segment, and the line segment being delivered to main leaf from sheet is from sheet line segment.
After fit, adopt following method to carry out coupling pre-service: when two line segment angles be less than 4 °, vertical range is not more than 5 pixels, and when two line segments have a lap, if line segment length difference is more than 50%, retain longer line segment, remove shorter line segment; Otherwise do line segment to merge, get the equal straight line of distance two line segments, and using two line segments farthest two-end-point as the line segment end points after merging.
Step 3, determine the sets of line segments doubtful of the same name that main leaf line segment is corresponding.
Travel through all main leaf line segments, when a certain meet following three conditions from sheet line segment simultaneously time, this be the line segment doubtful of the same name of corresponding main leaf line segment from sheet line segment:
A, with main leaf line segment angle be less than 5 °;
B, with main leaf line segment overlap length account for more than 20% of main leaf line segment length;
C, be not more than 10 pixels to the vertical range of main leaf line segment.
Find all doubtful line segment of the same name corresponding with main leaf line segment from sheet line segment, just constitute the sets of line segments doubtful of the same name of this main leaf line segment.
The concrete grammar that the similarity of line segment lap gray scale that utilizes the present embodiment reduces doubtful sets of line segments scope of the same name is as follows:
Extend out along vertical main leaf line segment direction is symmetrical, formed and extend out rectangle, described in calculating, extend out the remote sensing image average gray M in territory, rectangular foot-print; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the average gray M ＇ that the doubtful line segment of the same name of every bar extends out the remote sensing image in territory, rectangular foot-print in the same way; When M/M ＇ be greater than 2 or be less than 1/2 time, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
The present embodiment utilizes line segment left and right grey scale signal than the scope reducing doubtful sets of line segments of the same name further, and concrete grammar is as follows:
Extend out along vertical main leaf line segment direction is symmetrical, form two rectangles, according to the gray scale of the remote sensing image in territory, two rectangular foot-print, record the side that gray-scale value is high; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the remote sensing image left and right gray scale in the territory, two rectangular foot-print of the doubtful line segment of the same name of every bar in the same way, if the side that doubtful line segment gray-scale value of the same name is a high side position high from main leaf line segment gray-scale value is different, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
The sets of line segments optimization doubtful of the same name that the present embodiment also utilizes Kmeans clustering algorithm corresponding to main leaf line segment, further reduce the scope of doubtful sets of line segments of the same name, concrete grammar is:
Calculate the vertical range of main leaf line segment to its doubtful line segment of the same name, Kmeans clustering algorithm is utilized to be divided into two classes to doubtful line segment of the same name according to above-mentioned vertical range, the line segment doubtful of the same name that vertical range is larger is classified as a class automatically, the line segment doubtful of the same name that vertical range value is less is classified as another kind of automatically, rejects line segments doubtful of the same name all in the large class of vertical range.
Step 4, to determine from sets of line segments doubtful of the same name corresponding to sheet line segment.
Refer step 3, travel through all from sheet line segment, find in main leaf line segment with from all doubtful line segment of the same name corresponding to sheet line segment, generate this sets of line segments doubtful of the same name from sheet line segment, and utilize line segment left and right grey scale signal ratio and Kmeans clustering algorithm to optimize doubtful sets of line segments of the same name, thus reduce the scope of doubtful sets of line segments of the same name.
Step 5, rejecting error hiding.
Compare from the corresponding relation between line segment and its doubtful line segment of the same name in corresponding relation between the main leaf line segment that step 3 is obtained and its doubtful line segment of the same name and step 4, when main leaf line segment L with step 3, step 4, all there is corresponding relation from sheet line segment L ＇, then main leaf line segment L is with from sheet line segment L ＇, the match is successful, otherwise rejects main leaf line segment L and the matching relationship from sheet line segment L ＇.
According to above step, respectively as shown in Figure 6 and Figure 7, number identical line segment is the coupling line segment pair that the present invention obtains to the effect of line segment Auto-matching.Associating Fig. 6 and Fig. 7, the line segment of coupling is 45 right to having, and wherein correct coupling is 39 right, and error hiding 6 is right, and accuracy reaches 87%.Experimental result shows, the present invention can be effectively applied to the line match process of general significance.
Lines matching method based on affine invariants and homography matrix of the present invention is not limited to the concrete technical scheme described in above-described embodiment, and all employings are equal to replaces the protection domain that the technical scheme formed is application claims.
Claims (7)
1., based on a lines matching method for affine invariants and homography matrix, comprise the following steps:
Step 1, obtain optimum homography matrix between remote sensing image pair---based on the RANSAC iterative algorithm of affine invariants coupling, obtain the optimum homography matrix between remote sensing image pair;
Step 2, transmit remote sensing image between line segment---the line segment extracting remote sensing image centering respectively obtains line chart, and the optimum homography matrix obtained with step 1 is constraint, with one of them line chart for main leaf, another line chart is from sheet, to be delivered on main leaf from the line segment on sheet, original line segment wherein on main leaf is main leaf line segment, and the line segment being delivered to main leaf from sheet is from sheet line segment;
Step 3, determine the sets of line segments doubtful of the same name that main leaf line segment is corresponding---travel through all main leaf line segments, according to the direction between master and slave line segment, Distance geometry degree of overlapping, find all doubtful line segment of the same name corresponding with main leaf line segment from sheet line segment, form the sets of line segments doubtful of the same name of this main leaf line segment;
Step 4, determine from sets of line segments doubtful of the same name corresponding to sheet line segment---travel through all from sheet line segment, according to the direction between master and slave line segment, Distance geometry degree of overlapping, find in main leaf line segment with from all doubtful line segment of the same name corresponding to sheet line segment, form this sets of line segments doubtful of the same name from sheet line segment;
Step 5, rejecting error hiding---compare from the corresponding relation between sheet line segment and its doubtful line segment of the same name in the corresponding relation between the main leaf line segment that step 3 is obtained and its doubtful line segment of the same name and step 4, when main leaf line segment L with step 3, step 4, all there is corresponding relation from sheet line segment L ＇, then described main leaf line segment L with the match is successful from sheet line segment L ＇, otherwise rejects main leaf line segment L and the matching relationship from sheet line segment L ＇;
Based on the RANSAC iterative algorithm of affine invariants coupling in described step 1, concrete grammar is as follows:
Step I, use MSER extract operator, SIFT as feature interpretation operator as affine invariants, carry out the feature extraction of two width images, generating feature vector;
Step II, do distance operation between the proper vector of two subpictures, the SIFT feature vector obtaining coupling is right;
Step III, use RANSAC method are vectorial to processing to the SIFT feature of coupling, the input parameter of RANSAC method is distance threshold, the geometric model of input is the homography matrix of unknown parameters, obtain after process intra-office SIFT feature vector to and the design parameter of homography matrix, when performing this step first RANSAC method distance threshold span for (0,1];
The homography matrix that step IV, basis estimate is to remote sensing image to mating, and calculate matching accuracy rate, described matching accuracy rate is correct number of pairs and the ratio mated sum;
Step V, progressively increase the distance threshold of RANSAC method, and repeat step III to step IV until matching accuracy rate declines from 100%, matching accuracy rate be 100% homography matrix corresponding to maximal distance threshold be exactly optimum homography matrix.
2. the lines matching method based on affine invariants and homography matrix according to claim 1, it is characterized in that: in step IV, if the degree of overlapping of the characteristic area of the same name that remote sensing image is right is greater than 50%, then one_to_one corresponding both thinking, namely this coupling is correct coupling.
3. the lines matching method based on affine invariants and homography matrix according to claim 1, it is characterized in that: in described step II, during coupling SIFT feature point, when minimal characteristic vector distance is greater than 0.6 with the ratio of time minimal characteristic vector distance, proper vector is apart from the proper vector pair of that minimum a pair proper vector as coupling.
4. the lines matching method based on affine invariants and homography matrix according to claim 1, is characterized in that, after step 2 completes, step 3 carries out pre-service to the line segment of remote sensing image before performing, and concrete grammar is as follows:
When two line segment angles be less than 4 °, vertical range is not more than 5 pixels, and two line segments have lap: when line segment length difference is more than 50%, retain line segment longer in two line segments, remove line segment shorter in two line segments; Otherwise do line segment to merge, get the equal straight line of distance two line segments, and using two line segments farthest two-end-point as the line segment end points after merging.
5. the lines matching method based on affine invariants and homography matrix according to claim 1, is characterized in that, in step 3, when a certain meet following three conditions from sheet line segment simultaneously time, this be the line segment doubtful of the same name of corresponding main leaf line segment from sheet line segment:
A, with main leaf line segment angle be less than 5 °;
B, with main leaf line segment overlap length account for more than 20% of main leaf line segment length;
C, be not more than 10 pixels to the vertical range of main leaf line segment.
6. the lines matching method based on affine invariants and homography matrix according to claim 5, is characterized in that: utilize the similarity of line segment lap gray scale to reduce the scope of sets of line segments doubtful of the same name corresponding to main leaf line segment, concrete grammar is as follows:
Extend out along vertical main leaf line segment direction is symmetrical, formed and extend out rectangle, described in calculating, extend out the remote sensing image average gray M in territory, rectangular foot-print; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the average gray M ＇ that the doubtful line segment of the same name of every bar extends out the remote sensing image in territory, rectangular foot-print in the same way; When M/M ＇ be greater than 2 or be less than 1/2 time, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
7. the lines matching method based on affine invariants and homography matrix according to claim 5, it is characterized in that, utilize line segment left and right grey scale signal than the scope reducing sets of line segments doubtful of the same name corresponding to main leaf line segment, concrete grammar is as follows:
Extend out along vertical main leaf line segment direction is symmetrical, form two rectangles, according to the gray scale of the remote sensing image in territory, two rectangular foot-print, record the side that gray-scale value is high; In the sets of line segments doubtful of the same name that this main leaf line segment is corresponding, calculate the remote sensing image left and right gray scale in the territory, two rectangular foot-print of the doubtful line segment of the same name of every bar in the same way, if the side that doubtful line segment gray-scale value of the same name is a high side position high from main leaf line segment gray-scale value is different, this doubtful line segment of the same name is rejected from doubtful sets of line segments of the same name, otherwise retains this doubtful line segment of the same name.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210342566.3A CN102930525B (en) | 2012-09-14 | 2012-09-14 | Line matching method based on affine invariant feature and homography |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210342566.3A CN102930525B (en) | 2012-09-14 | 2012-09-14 | Line matching method based on affine invariant feature and homography |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102930525A CN102930525A (en) | 2013-02-13 |
| CN102930525B true CN102930525B (en) | 2015-04-15 |
Family
ID=47645315
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210342566.3A Expired - Fee Related CN102930525B (en) | 2012-09-14 | 2012-09-14 | Line matching method based on affine invariant feature and homography |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102930525B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103353984A (en) * | 2013-04-03 | 2013-10-16 | 西安交通大学 | Method for matching multiple image segments by using non geometric constraints |
| CN103679676A (en) * | 2013-12-02 | 2014-03-26 | 西北工业大学 | Quick unordered image stitching method based on multi-level word bag clustering |
| CN104616280B (en) * | 2014-11-26 | 2017-07-28 | 西安电子科技大学 | Method for registering images based on maximum stable extremal region and phase equalization |
| CN104392457B (en) * | 2014-12-11 | 2017-07-11 | 中国测绘科学研究院 | Incline the tie point automatic matching method and device of image |
| CN105005962B (en) * | 2015-08-14 | 2018-01-12 | 南京大学 | Islands and reefs Remote Sensing Image Matching method based on hierarchical screening strategy |
| CN105809678B (en) * | 2016-03-04 | 2018-08-31 | 中国民航大学 | A kind of line segment feature global registration method between two views under short base line condition |
| TWI618644B (en) * | 2016-05-06 | 2018-03-21 | 財團法人金屬工業研究發展中心 | Image overlay method |
| CN106529591A (en) * | 2016-11-07 | 2017-03-22 | 湖南源信光电科技有限公司 | Improved MSER image matching algorithm |
| CN110470295B (en) * | 2018-05-09 | 2022-09-30 | 北京智慧图科技有限责任公司 | Indoor walking navigation system and method based on AR positioning |
| CN108830797A (en) * | 2018-05-24 | 2018-11-16 | 桂林航天工业学院 | A kind of matching line segments method based on affine projection matrix model |
| CN110874850A (en) * | 2018-09-04 | 2020-03-10 | 湖北智视科技有限公司 | Real-time unilateral grid feature registration method oriented to target positioning |
| CN109919958B (en) * | 2019-01-14 | 2023-03-28 | 桂林航天工业学院 | Multi-constraint line segment extraction method based on multi-scale image space |
| CN110136159B (en) * | 2019-04-29 | 2023-03-31 | 辽宁工程技术大学 | Line segment extraction method for high-resolution remote sensing image |
| CN111582296B (en) * | 2019-12-20 | 2021-02-05 | 珠海大横琴科技发展有限公司 | Remote sensing image comprehensive matching method and device, electronic equipment and storage medium |
| CN112085771B (en) * | 2020-08-06 | 2023-12-05 | 深圳市优必选科技股份有限公司 | Image registration method, device, terminal equipment and computer readable storage medium |
| CN112163622B (en) * | 2020-09-30 | 2022-07-05 | 山东建筑大学 | Global and local fusion constrained aviation wide-baseline stereopair line segment matching method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101464948A (en) * | 2009-01-14 | 2009-06-24 | 北京航空航天大学 | Object identification method for affine constant moment based on key point |
| CN101650778A (en) * | 2009-07-28 | 2010-02-17 | 复旦大学 | Invariance identification method based on characteristic point and homography matching |
-
2012
- 2012-09-14 CN CN201210342566.3A patent/CN102930525B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101464948A (en) * | 2009-01-14 | 2009-06-24 | 北京航空航天大学 | Object identification method for affine constant moment based on key point |
| CN101650778A (en) * | 2009-07-28 | 2010-02-17 | 复旦大学 | Invariance identification method based on characteristic point and homography matching |
Non-Patent Citations (4)
| Title |
|---|
| 《遥感影像仿射不变特征匹配的自动优化》;程亮等;《武汉大学学报信息科学版》;20090430;第34卷(第4期);第418-419页第1.1节，第419页的1.2.1-1.2.2节，第420页左栏第7段及图1 * |
| 《面向宽基线立体影像匹配的高质量仿射不变特征提取方法》;程亮等;《测绘学报》;20080228;第37卷(第1期);第77-82页 * |
| Liang Cheng et al.《A New Method for Remote Sensing Image Matching by Integrating Affine Invariant Feature Extraction and RANSAC》.《3rd International Congress on Image and Signal Processing》.2010,第1605-1609页. * |
| Liang Cheng et al.《Remote sensing image matching by integrating affine invariant feature extraction and RANSAC》.《Computers and Electrical Engineering》.2012,第38卷(第4期),第1023-1032段. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102930525A (en) | 2013-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102930525B (en) | Line matching method based on affine invariant feature and homography | |
| CN111028277B (en) | SAR and optical remote sensing image registration method based on pseudo-twin convolution neural network | |
| CN102411778B (en) | Automatic registration method of airborne laser point cloud and aerial image | |
| CN105184801A (en) | Optical and SAR image high-precision registration method based on multilevel strategy | |
| CN101216895B (en) | An automatic extracting method for ellipse image features in complex background images | |
| CN102750537B (en) | Automatic registering method of high accuracy images | |
| CN103604417B (en) | The multi-view images bi-directional matching strategy that object space is information constrained | |
| CN103093459B (en) | Utilize the method that airborne LiDAR point cloud data assisted image mates | |
| CN104036480B (en) | Quick elimination Mismatching point method based on surf algorithm | |
| CN103727930A (en) | Edge-matching-based relative pose calibration method of laser range finder and camera | |
| CN107862735B (en) | RGBD three-dimensional scene reconstruction method based on structural information | |
| CN104077760A (en) | Rapid splicing system for aerial photogrammetry and implementing method thereof | |
| CN113658337B (en) | Multi-mode odometer method based on rut lines | |
| CN104121902A (en) | Implementation method of indoor robot visual odometer based on Xtion camera | |
| CN104408772A (en) | Grid projection-based three-dimensional reconstructing method for free-form surface | |
| CN104268866A (en) | Video sequence registering method based on combination of motion information and background information | |
| CN105513094A (en) | Stereo vision tracking method and stereo vision tracking system based on 3D Delaunay triangulation | |
| CN102750696A (en) | Affine invariant feature and coastline constraint-based automatic coastal zone remote-sensing image registration method | |
| CN104657986A (en) | Quasi-dense matching extension method based on subspace fusion and consistency constraint | |
| CN104167000A (en) | Affine-invariant wide-baseline image dense matching method | |
| CN103632376A (en) | Method for suppressing partial occlusion of vehicles by aid of double-level frames | |
| Hofer et al. | Line-based 3D reconstruction of wiry objects | |
| CN105654479A (en) | Multispectral image registering method and multispectral image registering device | |
| Li et al. | A new 3D LIDAR-based lane markings recognition approach | |
| CN103578105A (en) | Method for multisource different-type image registration based on region features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150415 Termination date: 20150914 |
|
| EXPY | Termination of patent right or utility model |