CN110781903B - Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint - Google Patents

Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint Download PDF

Info

Publication number
CN110781903B
CN110781903B CN201910969666.0A CN201910969666A CN110781903B CN 110781903 B CN110781903 B CN 110781903B CN 201910969666 A CN201910969666 A CN 201910969666A CN 110781903 B CN110781903 B CN 110781903B
Authority
CN
China
Prior art keywords
image
target image
grid
gridded
formula
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
Application number
CN201910969666.0A
Other languages
Chinese (zh)
Other versions
CN110781903A (en
Inventor
许权
罗林波
陈珺
龚文平
王勇
韩涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China University of Geosciences
Original Assignee
China University of Geosciences
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China University of Geosciences filed Critical China University of Geosciences
Priority to CN201910969666.0A priority Critical patent/CN110781903B/en
Publication of CN110781903A publication Critical patent/CN110781903A/en
Application granted granted Critical
Publication of CN110781903B publication Critical patent/CN110781903B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
    • G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
    • 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
    • G06T5/00—Image enhancement or restoration
    • G06T5/80—Geometric correction
    • 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
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/32—Indexing scheme for image data processing or generation, in general involving image mosaicing

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Image Processing (AREA)

Abstract

The invention provides an unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint, which specifically comprises the following steps: respectively extracting the feature points of the reference image I and the target image I' and the corresponding relation thereof by adopting an SIFT feature point extraction algorithm; gridding the reference image and the target image; obtaining a local homography matrix of the reference image and the target image after gridding by adopting a local DLT method; performing grid optimization by using a minimum energy function to align the grid target image and the reference image so as to eliminate ghost images in the overlapping area and obtain a primary spliced image; taking the ground as a landmark, and adopting global similarity coplanarity constraint on the preliminary spliced image to eliminate projection distortion of a non-overlapped part to obtain a final spliced image; the invention has the beneficial effects that: the characteristics of the images shot by the unmanned aerial vehicle are combined, the characteristics of the images of the unmanned aerial vehicle can be well matched, accurate registration is realized, and an ideal splicing effect is achieved.

Description

Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint
Technical Field
The invention relates to the field of image processing, in particular to an unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint.
Background
The unmanned aerial vehicle has the functions of automatic take-off and landing, automatic driving, automatic navigation, automatic rapid and accurate positioning, automatic information acquisition and transmission and the like. Is particularly suitable for replacing human beings to complete tasks under difficult, severe or extreme environments. Unmanned aerial vehicles have wide application in military surveying and mapping, aerospace and commercial fields. However, due to the limitation of a single viewing angle, it is difficult to cover the entire target area. To obtain a complete scene of the desired object, multiple images need to be stitched together. Image stitching is a technique that combines overlapping regions of multiple images to form a panoramic image.
For drone images, it has some features. Unmanned aerial vehicle is small, receives external influence easily at the flight in-process. When the unmanned aerial vehicle slightly shakes, the camera cannot be adjusted in time, and the acquired images are influenced. The most important issue is image parallax. Furthermore, there are different depth differences in the image due to the undulations of the ground. When the method is adopted to splice images of the unmanned aerial vehicle, due to the influence of the characteristics, double images and distortion can be generated.
Disclosure of Invention
The invention provides an unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint aiming at ghosting and distortion in the unmanned aerial vehicle image splicing process.
The invention solves the technical problem, and the adopted unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint comprises the following steps:
s101: extracting the feature points of the reference image I and the target image I 'by adopting an SIFT feature matching extraction algorithm, and performing feature matching to obtain the corresponding relation of the feature points of the reference image I and the target image I';
s102: gridding the reference image I and the target image I 'by utilizing the corresponding relation of the characteristic points of the reference image I and the target image I' to obtain a gridded reference image and a gridded target image;
s103: solving a local homography matrix of the gridded target image by adopting a local DLT method to obtain a local homography matrix corresponding to each grid in the gridded target image;
s104: according to the local homography matrix of each grid of the gridded target image and the vertex of each grid corresponding to the local homography matrix, guiding the deformation of the gridded target image by adopting a minimum energy function, aligning the gridded target image with the gridded reference image, and finishing grid optimization so as to eliminate double images of an overlapping area between the gridded target image and the gridded reference image and obtain a primary spliced image;
s105: and (3) taking the ground as a landmark, and eliminating the projection distortion of the non-overlapped part in the preliminary spliced image by utilizing global similarity coplanarity constraint to obtain a final spliced image.
Further, step S103 specifically includes the following steps: any pixel point x of the gridded target image*Matched to the gridded reference image and expressed by formula (1):
Figure BDA0002231638290000021
in the formula (1)
Figure BDA0002231638290000022
And
Figure BDA0002231638290000023
are respectively x*And y*Is represented by the same scale transformation, H*A local homography matrix corresponding to each grid of the gridded target image; h*Represented by the formula (2) h*Rebuilding to obtain;
Figure BDA0002231638290000024
in the formula (2), xiRepresenting each grid vertex of the gridded target image, delta is a preset scaling parameter,
Figure BDA0002231638290000025
is a scalar weight function, aiTwo rows of a matrix of two matching points, h is a preset estimate, | | aih is algebraic error, and h is solved through weight estimation*And obtaining a local homography matrix H after reconstruction*Correspondingly, each grid has a local sheetStress matrix HjJ is 1,2.. k, k denotes the number of meshes of the gridded target image.
5. Further, the expression of the minimized energy function in step S104 is as follows (3):
E＝Ea+αEr+βEs (3)
in the formula (3), EaTo align items, ErFor locally rotating terms, EsAnd alpha and beta are respectively preset weighted values for the local scale items.
Further, the alignment term E of the minimized energy function in step S104aThe expression is shown in formula (4):
Figure BDA0002231638290000031
in the formula (4), i is 1,2, k, k is the number of meshes of the gridded target image;
Figure BDA0002231638290000032
is a local homography matrix H corresponding to each grid in the target image after griddingjProjecting the gridded target image to the projection position of the gridded reference image;
Figure BDA0002231638290000033
is the corresponding grid on the target image after the grid optimization; w ═ w1，w2，w3，w4]TIs a preset bilinear weight;
Figure BDA0002231638290000034
to represent
Figure BDA0002231638290000035
Four mesh vertex coordinate sets are located;
dividing each uniform grid into two triangles, and defining the triangle in one grid as delta V1V2V3Then the transformed partRotation term ErIs represented by formula (5):
Figure BDA0002231638290000036
in the formula (5), the reaction mixture is,
Figure BDA0002231638290000037
tau is a weight coefficient for measuring the significance of the triangle, and mu and nu are in a local coordination system defined by other two vertexes
Figure BDA0002231638290000038
The coordinates of (a);
given a quadrilateral V1V2V3V4Then the local scale term EsIs represented by formula (6):
Figure BDA0002231638290000039
in the formula (6), EsAnd (V) is an energy term of the grid.
Further, step S105 specifically includes the following steps: and (3) taking the ground as a landmark, performing weighted updating on the non-overlapped part belonging to the target image in the preliminary spliced image, wherein the expression is shown as (7):
H`i＝ιHi+κS (7)
in the formula (7), HiIs the local homography matrix H' of the ith grid of the target image in the preliminary splicing imageiIs an alternative to the updated homography matrix, S is the global similarity transformation of the ground plane, and iota and kappa are preset weight coefficients, where iota + kappa is 1;
carrying out weighted updating on the non-overlapped part belonging to the reference image in the preliminary splicing image, wherein the weighted updating is shown as a formula (8);
Figure BDA0002231638290000041
in the formula (8), TiIs the local homography matrix of the reference image after the ith grid of the reference image in the preliminary splicing image is updated, and utilizes the local homography matrix T ″iEliminating projection distortion of non-overlapped part belonging to target image in preliminary splicing image, and utilizing the local homography matrix H ″iAnd eliminating projection distortion of the non-overlapping part of the reference image in the preliminary spliced image to obtain a final spliced image.
Further, after the final image splicing is completed, a groudtuth method is established, and the distortion degree of the reference spliced image after the grid optimization and global similarity coplanarity constraint is adopted is quantitatively evaluated.
And (3) estimating the distortion degree by using root mean square error quantitative analysis, as shown in an expression (9):
Figure BDA0002231638290000042
in the formula (9), P and P' are twisted portions in the group, N is the number of pixels of the twisted portion, and P isiAnd p' are usediIs the pixel of the distorted part in groudtuth, the smaller the RMSE value, the better.
The technical scheme provided by the invention has the beneficial effects that: the characteristics of the images shot by the unmanned aerial vehicle are combined, the characteristics of the images of the unmanned aerial vehicle can be well matched, accurate registration is realized, and an ideal splicing effect is achieved.
Drawings
FIG. 1 is a flowchart of an unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint according to the present invention;
FIG. 2 is a diagram illustrating the results of a comparative experiment of an unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint according to an embodiment of the present invention;
fig. 3 is a diagram of an application result of unmanned aerial vehicle image stitching based on mesh optimization and global similarity constraint in an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be further described with reference to the accompanying drawings.
Referring to fig. 1, an embodiment of the present invention provides an unmanned aerial vehicle image stitching method based on mesh optimization and global similarity constraint:
s101: extracting the feature points of the reference image I and the target image I 'by adopting an SIFT feature matching extraction algorithm, and performing feature matching to obtain the corresponding relation of the feature points of the reference image I and the target image I';
s102: gridding the reference image I and the target image I 'by utilizing the corresponding relation of the characteristic points of the reference image I and the target image I' to obtain a gridded reference image and a gridded target image;
s103: solving a local homography matrix of the gridded target image by adopting a local DLT method to obtain a local homography matrix corresponding to each grid in the gridded target image;
s104: according to the local homography matrix of each grid of the gridded target image and the vertex of each grid corresponding to the local homography matrix, guiding the deformation of the gridded target image by adopting a minimum energy function, aligning the gridded target image with the gridded reference image, and finishing grid optimization so as to eliminate double images of an overlapping area between the gridded target image and the gridded reference image and obtain a primary spliced image;
s105: and (3) taking the ground as a landmark, and eliminating the projection distortion of the non-overlapped part in the preliminary spliced image by utilizing global similarity coplanarity constraint to obtain a final spliced image.
Step S103 is specifically as follows: any pixel point x of the gridded target image*Matched to the gridded reference image and expressed by formula (1):
Figure BDA0002231638290000051
in the formula (1)
Figure BDA0002231638290000052
And
Figure BDA0002231638290000053
are respectively x*And y*Is represented by the same scale transformation, H*A local homography matrix corresponding to each grid of the gridded target image; h*Represented by the formula (2) h*Rebuilding to obtain;
Figure BDA0002231638290000054
in the formula (2), xiRepresenting each grid vertex of the gridded target image, delta is a preset scaling parameter,
Figure BDA0002231638290000055
is a scalar weight function, aiTwo rows of a matrix of two matching points, h is a preset estimate, | | aih is algebraic error, and h is solved through weight estimation*And obtaining a local homography matrix H after reconstruction*Correspondingly, each grid has a local homography matrix HjJ is 1,2.. k, k denotes the number of meshes of the gridded target image.
The expression of the minimized energy function in step S104 is as follows (3):
E＝Ea+αEr+βEs (3)
in the formula (3), EaTo align items, ErFor locally rotating terms, EsAnd alpha and beta are respectively preset weighted values for the local scale items.
The alignment term E of the minimized energy function in step S104aThe expression is shown in formula (4):
Figure BDA0002231638290000061
in the formula (4), i is 1,2, k, k is the gridded target imageThe number of grids;
Figure BDA0002231638290000062
is a local homography matrix H corresponding to each grid in the target image after griddingjProjecting the gridded target image to the projection position of the gridded reference image;
Figure BDA0002231638290000063
is the corresponding grid on the target image after the grid optimization; w ═ w1，w2，w3，w4]TIs a preset bilinear weight;
Figure BDA0002231638290000064
to represent
Figure BDA0002231638290000065
Four mesh vertex coordinate sets are located;
dividing each uniform grid into two triangles, and defining the triangle in one grid as delta V1V2V3Then the transformed local rotation term ErIs represented by formula (5):
Figure BDA0002231638290000066
in the formula (5), the reaction mixture is,
Figure BDA0002231638290000067
tau is a weight coefficient for measuring the significance of the triangle, and mu and nu are in a local coordination system defined by other two vertexes
Figure BDA0002231638290000068
The coordinates of (a);
given a quadrilateral V1V2V3V4Then the local scale term EsIs represented by formula (6):
Figure BDA0002231638290000069
in the formula (6), EsAnd (V) is an energy term of the grid.
Step S105 is specifically as follows: and (3) taking the ground as a landmark, performing weighted updating on the non-overlapped part belonging to the target image in the preliminary spliced image, wherein the expression is shown as (7):
H`i＝ιHi+κS (7)
in the formula (7), HiIs the local homography matrix H' of the ith grid of the target image in the preliminary splicing imageiIs an alternative to the updated homography matrix, S is the global similarity transformation of the ground plane, and iota and kappa are preset weight coefficients, where iota + kappa is 1;
carrying out weighted updating on the non-overlapped part belonging to the reference image in the preliminary splicing image, wherein the weighted updating is shown as a formula (8);
Figure BDA0002231638290000071
in the formula (8), TiIs the local homography matrix of the reference image after the ith grid of the reference image in the preliminary splicing image is updated, and utilizes the local homography matrix T ″iEliminating projection distortion of non-overlapped part belonging to target image in preliminary splicing image, and utilizing the local homography matrix H ″iAnd eliminating projection distortion of the non-overlapping part of the reference image in the preliminary spliced image to obtain a final spliced image.
And after the final image splicing is finished, creating a grountruth method, and quantitatively evaluating the distortion degree of the reference spliced image after the grid optimization and global similar coplanarity constraints are adopted.
And (3) estimating the distortion degree by using root mean square error quantitative analysis, as shown in an expression (9):
Figure BDA0002231638290000072
in the formula (9), P and P' are twisted portions in the group, N is the number of pixels of the twisted portion, and P isiAnd p' are usediIs the pixel of the distorted part in groudtuth, the smaller the RMSE value, the better.
TABLE 1 comparison of the RMSE values for the methods
Autotitch process APAP method The patented method
RMSE 39.15 38.86 29.91
Referring to fig. 2, fig. 2 is a diagram illustrating a comparison test result of an unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint according to an embodiment of the present invention. FIG. 2(a) is two images to be stitched; FIG. 2(b) is the result of stitching with global homography; FIG. 2(c) shows the result of splicing by APAP; FIG. 2(d) shows the splicing result of the present invention; the last image of fig. 2(b), (c), (d) highlights the distorted part of the image, and the second, third and fourth images highlight the ghost part of the image.
Referring to fig. 3, fig. 3 is a diagram illustrating an application result of image stitching of an unmanned aerial vehicle based on mesh optimization and global similarity constraint according to an embodiment of the present invention; FIGS. 3(a), (b) are the reference image and the target image, respectively; deforming fig. 3(b) to fig. 3(c) by mesh optimization; and (3) converting the images (a) and (c) in the images 3 into images (d) and (f) in the images 3 by adopting the global similarity coplanarity constraint to eliminate the distortion of the non-overlapped area. And finally, splicing the images (d) and (f) in the figure 3 to obtain the image (e) in the figure 3.
The technical scheme provided by the invention has the beneficial effects that: the characteristics of the images shot by the unmanned aerial vehicle are combined, the characteristics of the images of the unmanned aerial vehicle can be well matched, accurate registration is realized, and an ideal splicing effect is achieved.
In this document, the terms front, back, upper and lower are used to define the positions of the devices in the drawings and the positions of the devices relative to each other, and are used for the sake of clarity and convenience in technical solution. It is to be understood that the use of the directional terms should not be taken to limit the scope of the claims.
The features of the embodiments and embodiments described herein above may be combined with each other without conflict.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (4)

1. An unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint is characterized by comprising the following steps: the method specifically comprises the following steps:
s101: extracting the feature points of the reference image I and the target image I 'by adopting an SIFT feature matching extraction algorithm, and performing feature matching to obtain the corresponding relation of the feature points of the reference image I and the target image I';
s102: gridding the reference image I and the target image I 'by utilizing the corresponding relation of the characteristic points of the reference image I and the target image I' to obtain a gridded reference image and a gridded target image;
s103: solving a local homography matrix of the gridded target image by adopting a local DLT method to obtain a local homography matrix corresponding to each grid in the gridded target image;
s104: according to the local homography matrix of each grid of the gridded target image and the vertex of each grid of the gridded target image, guiding the deformation of the gridded target image by adopting a minimum energy function, aligning the gridded target image with the gridded reference image, and finishing grid optimization to eliminate double images of an overlapping area between the gridded target image and the gridded reference image to obtain a primary spliced image;
s105: taking the ground as a landmark, and eliminating the projection distortion of the non-overlapped part in the preliminary spliced image by utilizing global similarity coplane constraint to obtain a final spliced image;
step S103 is specifically as follows: any pixel point x of the gridded target image*Matching to the gridded reference image, and expressing by formula (1):
Figure FDA0003417136600000011
in the formula (1)
Figure FDA0003417136600000012
And
Figure FDA0003417136600000013
are respectively x*And y*Is represented by the same scale transformation, H*A local homography matrix corresponding to each grid of the gridded target image; h*Represented by the formula (2) h*Rebuilding to obtain;
Figure FDA0003417136600000014
in the formula (2), xiRepresenting each grid vertex of the gridded target image, delta is a preset scaling parameter,
Figure FDA0003417136600000015
is aA scalar weight function, aiTwo rows of a matrix of two matching points, h is a preset estimate, | | aih is algebraic error, and h is solved through weight estimation*And obtaining a local homography matrix H after reconstruction*；
The expression of the minimized energy function in step S104 is as follows (3):
E＝Ea+αEr+βEs (3)
in the formula (3), EaTo align items, ErFor locally rotating terms, EsThe local scale items are alpha and beta which are respectively preset weighted values;
in step S104, the alignment term E of the minimized energy functionaThe expression is shown in formula (4):
Figure FDA0003417136600000021
in the formula (4), i is 1,2, k, k is the number of meshes of the gridded target image;
Figure FDA0003417136600000022
is a local homography matrix H corresponding to each grid in the target image after griddingjProjecting the gridded target image to the projection position of the gridded reference image;
Figure FDA0003417136600000023
is the corresponding grid on the target image after the grid optimization; w ═ w1，w2，w3，w4]TIs a preset bilinear weight;
Figure FDA0003417136600000024
to represent
Figure FDA0003417136600000025
Four mesh vertex coordinate sets are located;
will each beDividing a uniform grid into two triangles, and defining the triangle in a grid as delta V1V2V3Then the transformed local rotation term ErIs represented by formula (5):
Figure FDA0003417136600000026
in the formula (5), the reaction mixture is,
Figure FDA0003417136600000027
tau is a weight coefficient for measuring the significance of the triangle, and mu and nu are in a local coordination system defined by other two vertexes
Figure FDA0003417136600000028
The coordinates of (a);
given a quadrilateral V1V2V3V4Then the local scale term EsIs represented by formula (6):
Figure FDA0003417136600000029
in the formula (6), EsAnd (V) is an energy term of the grid.
2. The unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint of claim 1, wherein: step S105 is specifically as follows: and (3) taking the ground as a landmark, performing weighted updating on the non-overlapped part belonging to the target image in the preliminary spliced image, wherein the expression is shown as (7):
H`i＝ιHi+κS (7)
in the formula (7), HiIs the local homography matrix H' of the ith grid of the target image in the preliminary splicing imageiIs an alternative to the updated homography matrix, S is the global similarity transformation of the ground plane, and iota and kappa are preset weighting coefficients, where iota + kappa1；
Carrying out weighted updating on the non-overlapped part belonging to the reference image in the preliminary splicing image, wherein the weighted updating is shown as a formula (8);
Figure FDA0003417136600000031
in the formula (8), TiIs the local homography matrix of the reference image after the ith grid of the reference image in the preliminary splicing image is updated, and utilizes the local homography matrix T ″iEliminating projection distortion of non-overlapped part belonging to target image in preliminary splicing image, and utilizing the local homography matrix H ″iAnd eliminating projection distortion of the non-overlapping part of the reference image in the preliminary spliced image to obtain a final spliced image.
3. The unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint of claim 1, wherein: and after the final image splicing is finished, creating a grountruth method, and quantitatively evaluating the distortion degree of the reference spliced image after the grid optimization and global similar coplanarity constraints are adopted.
4. The unmanned aerial vehicle image stitching method based on grid optimization and global similarity constraint of claim 3, wherein: and (3) estimating the distortion degree by using root mean square error quantitative analysis, as shown in an expression (9):
Figure FDA0003417136600000032
in the formula (9), P and P' are twisted portions in the group, N is the number of pixels of the twisted portion, and P isiAnd p' are usediIs the pixel of the distorted part in groudtuth, the smaller the RMSE value, the better.
CN201910969666.0A 2019-10-12 2019-10-12 Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint Active CN110781903B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910969666.0A CN110781903B (en) 2019-10-12 2019-10-12 Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910969666.0A CN110781903B (en) 2019-10-12 2019-10-12 Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint

Publications (2)

Publication Number Publication Date
CN110781903A CN110781903A (en) 2020-02-11
CN110781903B true CN110781903B (en) 2022-04-01

Family

ID=69385187

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910969666.0A Active CN110781903B (en) 2019-10-12 2019-10-12 Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint

Country Status (1)

Country Link
CN (1) CN110781903B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111507904B (en) * 2020-04-22 2023-06-02 华中科技大学 Image stitching method and device for microscopic printing patterns
CN112435163B (en) * 2020-11-18 2022-10-18 大连理工大学 Unmanned aerial vehicle aerial image splicing method based on linear feature protection and grid optimization
CN112492197B (en) * 2020-11-18 2022-01-07 京东科技信息技术有限公司 Image processing method and related equipment
CN112950468A (en) * 2021-02-01 2021-06-11 咪咕文化科技有限公司 Image splicing method, electronic device and readable storage medium
CN112862683B (en) * 2021-02-07 2022-12-06 同济大学 Adjacent image splicing method based on elastic registration and grid optimization
CN112927276B (en) * 2021-03-10 2024-03-12 杭州海康威视数字技术股份有限公司 Image registration method, device, electronic equipment and storage medium
CN113873099B (en) * 2021-08-27 2024-04-12 山东信通电子股份有限公司 Video image stabilizing method, device and medium for power transmission channel

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110211043A (en) * 2019-05-11 2019-09-06 复旦大学 A kind of method for registering based on grid optimization for Panorama Mosaic
CN110223222A (en) * 2018-03-02 2019-09-10 株式会社理光 Image split-joint method, image splicing device and computer readable storage medium

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150131924A1 (en) * 2013-11-13 2015-05-14 Microsoft Corporation Creation of Rectangular Images from Input Images

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110223222A (en) * 2018-03-02 2019-09-10 株式会社理光 Image split-joint method, image splicing device and computer readable storage medium
CN110211043A (en) * 2019-05-11 2019-09-06 复旦大学 A kind of method for registering based on grid optimization for Panorama Mosaic

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Image stitching by line-guided local warping with global similarity constraint;Tian-Zhu Xiang et al.;《Pattern Recognition》;20181231;第482-487页 *

Also Published As

Publication number Publication date
CN110781903A (en) 2020-02-11

Similar Documents

Publication Publication Date Title
CN110781903B (en) Unmanned aerial vehicle image splicing method based on grid optimization and global similarity constraint
CN107578376B (en) Image splicing method based on feature point clustering four-way division and local transformation matrix
CN107564061B (en) Binocular vision mileage calculation method based on image gradient joint optimization
CN110966991B (en) Single unmanned aerial vehicle image positioning method without control point
CN110648398B (en) Real-time ortho image generation method and system based on unmanned aerial vehicle aerial data
CN107767440B (en) Cultural relic sequence image fine three-dimensional reconstruction method based on triangulation network interpolation and constraint
CN112102458A (en) Single-lens three-dimensional image reconstruction method based on laser radar point cloud data assistance
CN106447601B (en) Unmanned aerial vehicle remote sensing image splicing method based on projection-similarity transformation
US20140015924A1 (en) Rapid 3D Modeling
WO2016082797A1 (en) Method for modeling and registering three-dimensional scene structure based on single image
CN107679537A (en) A kind of texture-free spatial target posture algorithm for estimating based on profile point ORB characteristic matchings
CN110111250A (en) A kind of automatic panorama unmanned plane image split-joint method and device of robust
CN109115184B (en) Collaborative measurement method and system based on non-cooperative target
WO2021163157A1 (en) Collaborative 3d mapping and surface registration
CN112862683B (en) Adjacent image splicing method based on elastic registration and grid optimization
CN108759788B (en) Unmanned aerial vehicle image positioning and attitude determining method and unmanned aerial vehicle
CN112767245B (en) System and method for map splicing construction based on real-time video images of multiple unmanned aerial vehicles
Ma et al. An operational superresolution approach for multi-temporal and multi-angle remotely sensed imagery
CN113566793A (en) True orthoimage generation method and device based on unmanned aerial vehicle oblique image
CN110243390A (en) The determination method, apparatus and odometer of pose
CN116721231A (en) Three-dimensional reconstruction method and system for extensible scene based on unmanned aerial vehicle-mounted positioning
Chibunichev et al. Photogrammetry for archaeology: Collecting pieces together
CN108109118B (en) Aerial image geometric correction method without control points
CN107330934B (en) Low-dimensional cluster adjustment calculation method and system
CN115631317B (en) Tunnel lining ortho-image generation method and device, storage medium and terminal

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
EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20200211

Assignee: Kunming New World Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023420000121

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230522

Application publication date: 20200211

Assignee: Kunming Baosheng Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023420000122

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230522

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20200211

Assignee: New Yunteng Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023420000132

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230530

Application publication date: 20200211

Assignee: Yunnan Changdian Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023980035961

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230530

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20200211

Assignee: Yunnan Beian surveying and mapping Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023980036740

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230620

Application publication date: 20200211

Assignee: Yunnan Quanyan Technology Information Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023980036738

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230620

Application publication date: 20200211

Assignee: Yunnan Jiayi Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023980036739

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230621

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20200211

Assignee: Yunnan Youfu Information Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023420000224

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230706

EE01 Entry into force of recordation of patent licensing contract
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20200211

Assignee: Yunnan Shenma Technology Co.,Ltd.

Assignor: CHINA University OF GEOSCIENCES (WUHAN CITY)

Contract record no.: X2023420000238

Denomination of invention: Unmanned Aerial Vehicle Image Mosaic Method Based on Grid Optimization and Global Similarity Constraints

Granted publication date: 20220401

License type: Common License

Record date: 20230712