CN103489197B - A kind of urban aerial image corner feature matching process - Google Patents
A kind of urban aerial image corner feature matching process Download PDFInfo
- Publication number
- CN103489197B CN103489197B CN201310489252.0A CN201310489252A CN103489197B CN 103489197 B CN103489197 B CN 103489197B CN 201310489252 A CN201310489252 A CN 201310489252A CN 103489197 B CN103489197 B CN 103489197B
- Authority
- CN
- China
- Prior art keywords
- corner characteristics
- image
- corner
- same name
- window
- 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.)
- Active
Links
Landscapes
- Image Processing (AREA)
Abstract
A kind of urban aerial image corner feature matching process, comprises the initial internal and external orientation of left and right two aviation images of input and two aviation images, extracts the corner characteristics of buildings respectively on left and right two aviation images, for all corner characteristics extracted on left image, angle point core line corresponding on right image is calculated according to the initial internal and external orientation of two aviation images, core line on right image is expanded up and down certain limit and form the field of search on right image, for corner characteristics fallen in the field of search all on right image, corner characteristics initial matching of the same name is carried out one by one with corner characteristics to be matched on left image, obtain candidate's corner characteristics of the same name, then the image greyscale information in conjunction with corner characteristics present position carries out corner characteristics of the same name essence coupling, obtain the related coefficient between corner characteristics to be matched on left image and each candidate corner characteristics of the same name, get the maximum and corner characteristics being greater than preset correlation coefficient number threshold value of related coefficient to as final corner characteristics pair of the same name.
Description
Technical field
The invention belongs to Surveying Science and Technology field, relate to a kind of urban aerial image corner feature matching process, be mainly used in aviation image automatic triangulation, high precision DSM and automatically produce and the field such as true orthophoto production.
Background technology
Image Matching is the important step of aviation image aerotriangulation, is wherein mainly same place characteristic matching.Existing imaging point characteristic matching such as gray scale correlation coefficient matching method, SIFT mate scheduling algorithm and set up certain similarity measure according to the whole image texture information in certain limit centered by unique point and carry out Feature Points Matching of the same name, and the Image Matching for landform comparatively continuum has very good effect.But for large scale urban area aviation image, a large amount of existence due to buildings create a large amount of landform discontinuity zone.Enough ground match points cannot be obtained at the intensive place of buildings, and be often difficult to coupling acquisition same place because texture comparatively lacks at building roof place, thus cause Image Matching point skewness and same place lazy weight, be unfavorable for carrying out smoothly of follow-up aviation image area adjustment.Consider that buildings angle point is desirable unique point, but because buildings angle point is positioned at parallax discontinuity zone, imaging window on the image of left and right centered by angle point of the same name only has identical texture in side, roof, and because the image of height displacement causes texture inconsistent outside roof, therefore successfully cannot carry out the coupling of Corner Feature with traditional points correspondence algorithm.
Summary of the invention
The object of this invention is to provide a kind of urban aerial image corner feature matching process, to overcome in the aviation image matching process of urban area because of the homotopy mapping difficult problem that parallax is discontinuous, roof texture shortage etc. causes, sane same place information can be provided for urban area aviation image sky three.
Technical scheme of the present invention is a kind of urban aerial image corner feature matching process, comprises the following steps:
Step 1, inputs the initial internal and external orientation of left and right two aviation images and two aviation images;
Step 2, the corner characteristics of buildings is extracted respectively on left and right two aviation images, comprise and first on aviation image, extract linear feature, then according to extraction of straight line corner characteristics, each corner characteristics has adjacent end points by two and approximately perpendicular straight-line segment forms, the intersection point of these two straight-line segments is the angle point of corner characteristics, and according to from initial line to stop the counterclockwise angle in edge be less than the principle determination corner characteristics of 180 degree initial line and stop limit;
Step 3, all corner characteristics extracted on left image for step 2, carry out following matching operation one by one, until all corner characteristics couplings on left image are complete,
For a corner characteristics to be matched on left image, angle point core line corresponding on right image is calculated according to the initial internal and external orientation of two aviation images, core line on right image is expanded up and down certain limit and form the field of search on right image, for corner characteristics fallen in the field of search all on right image, carry out corner characteristics initial matching of the same name with corner characteristics to be matched on left image one by one, obtain candidate's corner characteristics of the same name;
For corner characteristics to be matched on left image and each candidate corner characteristics of the same name, image greyscale information respectively in conjunction with corner characteristics present position carries out corner characteristics of the same name essence coupling, obtain the related coefficient between corner characteristics to be matched on left image and each candidate corner characteristics of the same name, get the maximum and corner characteristics being greater than preset correlation coefficient number threshold value of related coefficient to as final corner characteristics pair of the same name.
And, the implementation of carrying out corner characteristics initial matching of the same name is, for corner characteristics corner characteristics to be matched on left image and right image fallen in the field of search, is limit of the same name with the termination limit of playing initial line or two corner characteristics of two corner characteristics, carry out following operation respectively
Employing rotation, translation and scale transformation model calculate the coordinate conversion parameter between two corner characteristics, under corner characteristics on right image is converted into corner characteristics coordinate system to be matched on left image, after coordinate transform, the limit of the same name of two corner characteristics overlaps, and then calculates the end points spacing of two corner characteristics;
Candidate's corner characteristics of the same name is obtained when minimum value is less than distance threshold in gained two end points spacing.
And, the implementation of carrying out corner characteristics of the same name essence coupling is, for candidate's corner characteristics of the same name on corner characteristics to be matched on left image and right image, respectively 1/4 respective valid window and 3/4 valid window are obtained to two corner characteristics, between 1/4 valid window of two corner characteristics, carry out matching primitives related coefficient C according to image greyscale information
1/4, between 3/4 valid window of two corner characteristics, carry out matching primitives related coefficient C according to image greyscale information
3/4, get related coefficient C
1/4and C
3/4in maximal value as the related coefficient between two corner characteristics.
And the implementation certain corner characteristics being obtained to corresponding 1/4 valid window and 3/4 valid window is as follows,
Step 4.1, with an initial line of corner characteristics for x-axis, with the angle point of corner characteristics for true origin O, set up rectangular coordinate system Oxy, and adopt affine Transform Model to be corrected in y-axis by isometric for an other limit of corner characteristics, calculate the conversion parameter between the front corner characteristics of conversion and new coordinate system Oxy;
Step 4.2, centered by the true origin O of rectangular coordinate system Oxy, imaging window is got according to pre-set dimension, the x-axis of rectangular coordinate system Oxy is parallel with two limits of imaging window respectively with y-axis, the coordinate of each pixel after coordinate transform on the image of corner characteristics place in imaging window is calculated successively according to step 4.1 gained conversion parameter, obtain grey scale pixel value according to this coordinate figure interpolation on this image, obtain the imaging window after correcting;
Step 4.3, for the imaging window after correction, is divided into 1/4 valid window and 3/4 valid window according to x-axis and y-axis by imaging window, only gets the part that imaging window is positioned at first quartile in 1/4 valid window; The part that imaging window is positioned at second and third and four-quadrant is only got in 3/4 valid window.
Advantage of the present invention carries out aviation image corner characteristics Auto-matching of the same name in conjunction with linear feature and image greyscale information, by extracting straight-line segment and then extract corner characteristics on aviation image, and on left and right image to be matched, carry out corner characteristics initial matching of the same name and corner characteristics of the same name essence coupling successively, obtain angle point information of the same name accurately.Because the present invention obtains after corner characteristics of the same name in coupling, can directly obtain line features of the same name, automatically can detect the discontinuous position of parallax for this reason and and then provide constraint condition for image dense Stereo Matching.Therefore the present invention to produce and the field such as true orthophoto production has a good application prospect automatically at urban area aviation image automatic triangulation, high precision DSM.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the embodiment of the present invention.
Fig. 2 is the corner detection schematic diagram of the embodiment of the present invention.
Fig. 3 be the embodiment of the present invention determine corner characteristics hunting zone of the same name schematic diagram based on core line.
Fig. 4 is the corner characteristics initial matching schematic diagram of the embodiment of the present invention.
Fig. 5 is the corner characteristics exact matching schematic diagram in conjunction with image greyscale information of the embodiment of the present invention.
Fig. 6 is that in the corner characteristics exact matching of the embodiment of the present invention, angle point 1/4 valid window and 3/4 valid window divide schematic diagram.
Embodiment
Technical solution of the present invention is described in detail below in conjunction with drawings and Examples.
Technical solution of the present invention can adopt computer software technology to realize automatic operational scheme.As shown in Figure 1, the flow process of embodiment comprises following steps:
Step 1, inputs the initial internal and external orientation of left and right two aviation images and two aviation images.
Step 2, extracts corner characteristics respectively on left and right two aviation images.When extracting corner characteristics, first on two aviation images, extract linear feature respectively, the linear feature major part extracted is buildings edge, and each corner characteristics has adjacent end points by two and approximately perpendicular straight-line segment forms, and the intersection point of these two straight-line segments is the angle point of corner characteristics.
For the sake of ease of implementation, the aviation image corner detection of embodiment is provided to realize as shown in Figure 2.Its step is as follows:
1. on aviation image, extract linear feature, suppose AB, CD, EF, GH are extracted buildings linear edge.
2. all straight-line segments that aviation image extracts are proceeded as follows: for straight line section (as AB in Fig. 2) any on image, certain threshold value (such as getting 80 °) is greater than with the acute angle of this straight-line segment and its a certain end points is with this end-point distances that nearest and distance is less than the straight line section of certain threshold value (such as getting 20 pixels) in the search of its any one end points (in as Fig. 2 B) place, successfully suppose that CD is the straight-line segment meeting search condition if search for, then these two adjacent straight-line segment AB-CD can form a corner characteristics.After extraction corner characteristics, by two Intersection of line segments, and replace corresponding straight-line segment end points with intersection point, if AB and CD intersects at M, then new corner characteristics is AM-MD(can be abbreviated as AMD).In like manner, if also search out a qualified straight-line segment (if HG) at its another terminal A place, another buildings corner characteristics can be formed.And according to from initial line to stop the counterclockwise angle in edge be less than the principle determination corner characteristics of 180 degree initial line and stop limit, as in corner characteristics AMD in Fig. 2, MD should be initial line, MA is termination limit.
Step 3, for all corner characteristics on left image, carries out following matching operation one by one, until all corner characteristics couplings on left image are complete,
For the corner characteristics to be matched of on left image, its angle point core line corresponding on right image is calculated according to the initial internal and external orientation of two aviation images, core line on right image is expanded up and down certain limit and form the field of search on right image, for corner characteristics fallen in the field of search all on right image, carry out corner characteristics initial matching of the same name with the corner characteristics to be matched on left image one by one, obtain multiple candidate corner characteristics of the same name;
For each candidate corner characteristics of the same name obtained, image greyscale information in conjunction with corner characteristics present position carries out corner characteristics of the same name essence coupling, calculate the related coefficient between corner characteristics to be matched and each corner characteristics of the same name, get the maximum and corner characteristics being greater than preset correlation coefficient number threshold value (such as getting 0.85) of related coefficient to as final corner characteristics pair of the same name, thus obtain that corner characteristics on left image is unique on right image, corner characteristics of the same name accurately.If related coefficient does not all reach threshold value, it fails to match.
For the sake of ease of implementation, what provide embodiment determines the corner characteristics field of search of the same name implementation as shown in Figure 3 based on core line, left and right two aviation images provided are respectively referred to as left image and right image, and based on left image in the enterprising line search coupling of right image, its step is as follows:
1. for the corner characteristics to be matched of on left image (as ABC in Fig. 3), according to the core line of the angle point (B) of initial internal and external orientation this corner characteristics calculated of two aviation images (as the corresponding epipolar line in Fig. 3 a) on right image (in as Fig. 3 b);
2. the core line on right image is expanded up and down certain limit (desirable expand 150 pixels up and down, as extended to b1 and b2 in Fig. 3), when specifically implementing, those skilled in the art can the width of sets itself spreading range;
3. all corner characteristics fallen into by angle point on right image within the scope of expansion nuclear line take out in order to participating in corner characteristics initial matching of the same name, as DEF, GHI, JKL in Fig. 3.
The present invention proposes further, employing rotation, translation and scale transformation model calculate the coordinate conversion parameter between two buildings corner characteristics, under buildings corner characteristics on right image being converted into the buildings corner characteristics coordinate system of left image, then the distance of the corresponding end points of two corner characteristics is calculated as corner characteristics similarity measure, carry out corner characteristics coupling of the same name when can there is rotation angle, different scale when between left and right image, corner characteristics initial matching method of the same name namely used has rotation and scale invariability.For the sake of ease of implementation, embodiment corner characteristics initial matching of the same name is provided to realize as shown in Figure 4.Its step is as follows:
1. for a corner characteristics (in Fig. 4 DEF) on the corner characteristics to be matched of on left image (as ABC in Fig. 4) and right image, using their angle point as a pair same place (as Fig. 4 mid point B and some E), select left image corner characteristics play initial line and right image corner characteristics play initial line as a pair limit of the same name (as BC and EF in Fig. 4), be a pair same place (as Fig. 4 mid point C and some F) with another of limit of the same name to end points;
2. the coordinate conversion parameter between rotation of the prior art, above two corner characteristics of Pan and Zoom transformation model calculating is adopted according to above two pairs of same places;
3. 2. walk according to the coordinate conversion parameter calculated, corner characteristics on right image is transformed in the coordinate system of corner characteristics on left image, as shown in Figure 4, after conversion, some E and some F overlaps with a B and some C respectively, and namely after coordinate transform, the initial line that rises of two corner characteristics overlaps.Then the distance between the end points that after calculating conversion, two corner characteristics are left, i.e. end points spacing (spacing of Fig. 4 mid point A and some D).
4. using the termination limit of conversion front left image corner characteristics and the termination limit of right image corner characteristics as limit of the same name (as BA and ED in Fig. 4), according to 1., 2., 3. in step make coordinate transform after the termination limit of two corner characteristics overlap, the end points spacing (namely Fig. 4 mid point E and some D overlaps with a B and some A respectively, and after calculating conversion, two corner characteristics end points spacing are spacing of a C and some F) of two corner characteristics after calculating coordinate change.And using the smallest end dot spacing of the minimum value in twice calculating as two corner characteristics, if the smallest end dot spacing of two corner characteristics be less than certain distance threshold (suggestion arrange threshold value be left image corner characteristics two line segment lengths and 1/3), then think that right image corner characteristics is candidate's corner characteristics of the same name of left image corner characteristics.
For the sake of ease of implementation, to provide in embodiment in conjunction with the corner characteristics exact matching implementation of image greyscale information as shown in Figure 5.Because most of buildings all has similar rectangular profile, and often there is dual edge phenomenon when extracting buildings edge because the low parapet in roof has one fixed width.Above situation all can cause on right image, occurring multiple corner characteristics meeting just matching condition when carrying out corner characteristics and just mating for a corner characteristics on left image, thus cannot correct judgment matching result.Therefore need to carry out corner characteristics essence coupling in conjunction with half-tone information near corner characteristics.And exist because elevation suddenlys change the parallax non-continuous event caused at buildings corner point, the correlation coefficient matching method of angle point cannot be carried out by the overall window of routine, thus the present invention propose further based on corner characteristics constraint angle point 1/4 valid window and 3/4 valid window matching algorithm.Embodiment performing step is as follows:
1. as shown in Figure 5, with an initial line BC of corner characteristics ABC for x-axis, with the angle point B of corner characteristics for true origin O, set up new rectangular coordinate system Oxy, and adopt affine Transform Model of the prior art to be corrected in y-axis by isometric for an other limit BA of corner characteristics, calculate the conversion parameter between the front corner characteristics ABC of conversion and new coordinate system Oxy;
2. centered by the true origin O of rectangular coordinate system Oxy, imaging window (x-axis and y-axis are parallel with two limits of imaging window respectively) is got according to pre-set dimension, the coordinate of each pixel after coordinate transform on the image of corner characteristics place in imaging window is calculated successively according to the conversion parameter of 1. middle calculating, obtain grey scale pixel value according to this coordinate figure interpolation on this image, thus obtain the imaging window after correcting; Those skilled in the art can sets itself window size, and such as pre-set dimension is 63 × 63 pixels;
3. for the imaging window (can be described as large window) after correction, according to x-axis and y-axis, window is divided into 1/4 valid window and 3/4 valid window (the image blocks participation Calculation of correlation factor as shown in Figure 6, in 1/4 valid window only represented by the white portion of fetch bit in first quartile; In 3/4 valid window only fetch bit in second and third, image blocks represented by white portion in four-quadrant participates in Calculation of correlation factor).
For the corner characteristics pair that corner characteristics to be matched and any one corresponding candidate corner characteristics of the same name are formed, respectively two corner characteristics are as above processed, obtain 1/4 respective valid window and 3/4 valid window, first mate between 1/4 valid window of two corner characteristics, calculate related coefficient C
1/4, then mate between 3/4 valid window of two corner characteristics, calculate related coefficient C
3/4, get maximal value in the related coefficient of twice calculating as the related coefficient between two corner characteristics.Calculating related coefficient gives image greyscale information and adopts existing gray scale correlation coefficient matching method method to realize, and it will not go into details in the present invention.
During concrete enforcement, those skilled in the art according to circumstances can preset the value of each threshold value voluntarily.
Specific embodiment described herein is only to the explanation for example of the present invention's spirit.Those skilled in the art can make various amendment or supplement or adopt similar mode to substitute to described specific embodiment, but can't depart from spirit of the present invention or surmount the scope that appended claims defines.
Claims (2)
1. a urban aerial image corner feature matching process, is characterized in that, comprises the following steps:
Step 1, inputs the initial internal and external orientation of left and right two aviation images and two aviation images;
Step 2, the corner characteristics of buildings is extracted respectively on left and right two aviation images, comprise and first on aviation image, extract linear feature, then according to extraction of straight line corner characteristics, each corner characteristics has adjacent end points by two and approximately perpendicular straight-line segment forms, the intersection point of these two straight-line segments is the angle point of corner characteristics, and according to from initial line to stop the counterclockwise angle in edge be less than the principle determination corner characteristics of 180 degree initial line and stop limit;
Step 3, all corner characteristics extracted on left image for step 2, carry out following matching operation one by one, until all corner characteristics couplings on left image are complete,
For a corner characteristics to be matched on left image, angle point core line corresponding on right image is calculated according to the initial internal and external orientation of two aviation images, core line on right image is expanded up and down certain limit and form the field of search on right image, for corner characteristics fallen in the field of search all on right image, carry out corner characteristics initial matching of the same name with corner characteristics to be matched on left image one by one, obtain candidate's corner characteristics of the same name;
Described implementation of carrying out corner characteristics initial matching of the same name is, for corner characteristics corner characteristics to be matched on left image and right image fallen in the field of search, be limit of the same name with the termination limit of playing initial line or two corner characteristics of two corner characteristics, carry out following operation respectively
Employing rotation, translation and scale transformation model calculate the coordinate conversion parameter between two corner characteristics, under corner characteristics on right image is converted into corner characteristics coordinate system to be matched on left image, after coordinate transform, the limit of the same name of two corner characteristics overlaps, and then calculates the end points spacing of two corner characteristics;
Candidate's corner characteristics of the same name is obtained when minimum value is less than distance threshold in gained two end points spacing;
For corner characteristics to be matched on left image and each candidate corner characteristics of the same name, image greyscale information respectively in conjunction with corner characteristics present position carries out corner characteristics of the same name essence coupling, obtain the related coefficient between corner characteristics to be matched on left image and each candidate corner characteristics of the same name, get the maximum and corner characteristics being greater than preset correlation coefficient number threshold value of related coefficient to as final corner characteristics pair of the same name;
Described implementation of carrying out corner characteristics of the same name essence coupling is, for candidate's corner characteristics of the same name on corner characteristics to be matched on left image and right image, respectively 1/4 respective valid window and 3/4 valid window are obtained to two corner characteristics, between 1/4 valid window of two corner characteristics, carry out matching primitives related coefficient C according to image greyscale information
1/4, between 3/4 valid window of two corner characteristics, carry out matching primitives related coefficient C according to image greyscale information
3/4, get related coefficient C
1/4and C
3/4in maximal value as the related coefficient between two corner characteristics.
2. urban aerial image corner feature matching process according to claim 1, is characterized in that: the implementation certain corner characteristics being obtained to corresponding 1/4 valid window and 3/4 valid window is as follows,
Step a, with an initial line of corner characteristics for x-axis, with the angle point of corner characteristics for true origin O, set up rectangular coordinate system Oxy, and adopt affine Transform Model to be corrected in y-axis by isometric for an other limit of corner characteristics, calculate the conversion parameter between the front corner characteristics of conversion and new coordinate system Oxy;
Step b, centered by the true origin O of rectangular coordinate system Oxy, imaging window is got according to pre-set dimension, the x-axis of rectangular coordinate system Oxy is parallel with two limits of imaging window respectively with y-axis, the coordinate of each pixel after coordinate transform on the image of corner characteristics place in imaging window is calculated successively according to step a gained conversion parameter, obtain grey scale pixel value according to this coordinate figure interpolation on this image, obtain the imaging window after correcting;
Step c, for the imaging window after correction, is divided into 1/4 valid window and 3/4 valid window according to x-axis and y-axis by imaging window, only gets the part that imaging window is positioned at first quartile in 1/4 valid window; The part that imaging window is positioned at second and third and four-quadrant is only got in 3/4 valid window.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310489252.0A CN103489197B (en) | 2013-10-18 | 2013-10-18 | A kind of urban aerial image corner feature matching process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310489252.0A CN103489197B (en) | 2013-10-18 | 2013-10-18 | A kind of urban aerial image corner feature matching process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103489197A CN103489197A (en) | 2014-01-01 |
| CN103489197B true CN103489197B (en) | 2016-03-09 |
Family
ID=49829395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310489252.0A Active CN103489197B (en) | 2013-10-18 | 2013-10-18 | A kind of urban aerial image corner feature matching process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103489197B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104535049B (en) * | 2014-12-19 | 2017-01-18 | 广东南方数码科技股份有限公司 | Aerial photograph non-stereoscopic collection mapping method |
| CN112953383A (en) * | 2017-08-30 | 2021-06-11 | 深圳市大疆创新科技有限公司 | Method and equipment for detecting photovoltaic panel and unmanned aerial vehicle |
| CN112163562B (en) * | 2020-10-23 | 2021-10-22 | 珠海大横琴科技发展有限公司 | Image overlapping area calculation method and device, electronic equipment and storage medium |
| CN113689481A (en) * | 2021-08-30 | 2021-11-23 | 南京信息工程大学 | Quick matching algorithm for homonymous cloud points based on medium-resolution images |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102411778A (en) * | 2011-07-28 | 2012-04-11 | 武汉大学 | Automatic registration method of airborne laser point cloud and aerial image |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8081798B2 (en) * | 2007-11-20 | 2011-12-20 | Lawrence Livermore National Security, Llc | Method and system for detecting polygon boundaries of structures in images as particle tracks through fields of corners and pixel gradients |
-
2013
- 2013-10-18 CN CN201310489252.0A patent/CN103489197B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102411778A (en) * | 2011-07-28 | 2012-04-11 | 武汉大学 | Automatic registration method of airborne laser point cloud and aerial image |
Non-Patent Citations (2)
| Title |
|---|
| 城区机载LiDAR数据与航空影像的自动配准;张永军 等;《遥感学报》;20120525;第587-595页 * |
| 基于面积特征的城区航空影像匹配方法;田岩 等;《红外与激光工程》;20021031;第31卷(第5期);第371-374页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103489197A (en) | 2014-01-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021233029A1 (en) | Simultaneous localization and mapping method, device, system and storage medium | |
| CN102411778B (en) | Automatic registration method of airborne laser point cloud and aerial image | |
| CN110689622B (en) | Synchronous positioning and composition method based on point cloud segmentation matching closed-loop correction | |
| CN102930525B (en) | Line matching method based on affine invariant feature and homography | |
| 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 | |
| CN105096386A (en) | Method for automatically generating geographic maps for large-range complex urban environment | |
| CN103489197B (en) | A kind of urban aerial image corner feature matching process | |
| CN102519481A (en) | Implementation method of binocular vision speedometer | |
| CN105160702A (en) | Stereoscopic image dense matching method and system based on LiDAR point cloud assistance | |
| CN106548173A (en) | A kind of improvement no-manned plane three-dimensional information getting method based on classification matching strategy | |
| CN104732577A (en) | Building texture extraction method based on UAV low-altitude aerial survey system | |
| CN110130987B (en) | Tunnel convergence deformation monitoring method based on image analysis | |
| CN116449384A (en) | Radar inertial tight coupling positioning mapping method based on solid-state laser radar | |
| Zhang et al. | Automatic registration of urban aerial imagery with airborne LiDAR data | |
| CN101556694B (en) | Matching method of rotating images | |
| CN105184786A (en) | Floating-point-based triangle characteristic description method | |
| KR20180015961A (en) | Method of estimating the location of object image-based and apparatus therefor | |
| CN112700486A (en) | Method and device for estimating depth of road lane line in image | |
| CN105225233A (en) | A kind of stereopsis dense Stereo Matching method and system expanded based on two classes | |
| CN107578429B (en) | Stereo image dense matching method based on dynamic programming and global cost accumulation path | |
| CN103473782B (en) | Based on the Least squares matching method of the vertical two-sided unit of object space | |
| CN110610503B (en) | Three-dimensional information recovery method for electric knife switch based on three-dimensional matching | |
| CN111197976A (en) | Three-dimensional reconstruction method considering multi-stage matching propagation of weak texture region | |
| CN111462195B (en) | Irregular angle direction cost aggregation path determination method based on dominant line constraint |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20180731 Address after: 150025 College Road, Harbin Limin Development Zone, Heilongjiang Province Patentee after: Heilongjiang Longfei aviation photography Co. Ltd. Address before: 430072 Wuhan University, Luojia mountain, Wuchang District, Wuhan, Hubei Patentee before: Wuhan University |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211202 Address after: 150000 No. 2, tiesan street, Jianbao Road, Nangang District, Harbin, Heilongjiang Patentee after: The second Institute of aerial survey and remote sensing, Ministry of natural resources Address before: 150025 College Road, Harbin Limin Development Zone, Heilongjiang Province Patentee before: HEILONGJIANG LONGFEI AVIATION PHOTOGRAPHY Co.,Ltd. |