CN108171787A - A kind of three-dimensional rebuilding method based on the detection of ORB features - Google Patents

A kind of three-dimensional rebuilding method based on the detection of ORB features Download PDF

Info

Publication number
CN108171787A
CN108171787A CN201711366005.6A CN201711366005A CN108171787A CN 108171787 A CN108171787 A CN 108171787A CN 201711366005 A CN201711366005 A CN 201711366005A CN 108171787 A CN108171787 A CN 108171787A
Authority
CN
China
Prior art keywords
point
camera
matrix
detection
characteristic point
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.)
Pending
Application number
CN201711366005.6A
Other languages
Chinese (zh)
Inventor
周祖鹏
张晓东
甘良棋
裴雨蒙
钟雪波
蒋开云
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.)
Guilin University of Electronic Technology
Original Assignee
Guilin University of Electronic Technology
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 Guilin University of Electronic Technology filed Critical Guilin University of Electronic Technology
Priority to CN201711366005.6A priority Critical patent/CN108171787A/en
Publication of CN108171787A publication Critical patent/CN108171787A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/40—Image enhancement or restoration using histogram techniques
    • 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
    • G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Geometry (AREA)
  • Computer Graphics (AREA)
  • Software Systems (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a kind of three-dimensional rebuilding methods based on the detection of ORB features, belong to feature detection techniques field, realtime graphic and pretreatment are obtained by the binocular camera being mounted on unmanned machine head, video camera is demarcated, it solves camera inside and outside parameter and distortion factor to correct pixel coordinate, image characteristic point is detected with the detection of ORB features, is matched using FLANN, the space coordinate of characteristic point is obtained, three-dimensional reconstruction is carried out to spatial discrete points using OpenGL.The present invention can accelerate system and perform speed, while have rotational invariance and the robustness of anti-noise jamming, greatly improve the precision of real-time three-dimensional reconstruction.

Description

A kind of three-dimensional rebuilding method based on the detection of ORB features
Technical field
The present invention relates to feature detection techniques field more particularly to a kind of three-dimensional rebuilding methods based on the detection of ORB features.
Background technology
The feature point extraction of image and detection are parts crucial in three-dimensional reconstruction, are affecting Feature Points Matching just True rate and final reconstructed results.With universal and development, unmanned plane automatic obstacle avoiding and the path planning of unmanned plane and utilization Unmanned plane is acquired digital image information progress real-time three-dimensional reconstruction etc. and has been to be concerned by more and more people and pay attention to.Current three It ties up image characteristic point detection in reconstruction technique and includes the matching based on gray scale, feature-based matching and based on office with matching algorithm The matching of constant description in portion, wherein the images match based on local invariant description is special including SIFT feature description, SURF Sign description, D-nets Feature Descriptors etc..Unmanned plane in shooting picture, the picture that was photographed easily by ambient and Itself interference, generates picture noise.This has a great impact to the result of three-dimensional reconstruction.
Invention content
In view of the deficiencies of the prior art, problem solved by the invention is the problem of real-time three-dimensional reconstruction precision is not high.
In order to solve the above technical problems, the technical solution adopted by the present invention is a kind of Three-dimensional Gravity detected based on ORB features Construction method obtains realtime graphic and pretreatment by the binocular camera that is mounted on unmanned machine head, to video camera into rower It is fixed, it solves camera inside and outside parameter and distortion factor and pixel coordinate is corrected, image characteristic point is detected with the detection of ORB features, It is matched using FLANN, the space coordinate of characteristic point is obtained, spatial discrete points are carried out with three-dimensional reconstruction, packet using OpenGL Include following steps：
Step 1:Binocular camera shooting system is mounted on unmanned machine head, establishes unmanned aerial vehicle base station system, is passed in real time Defeated image；
Step 2:The image of acquisition is pre-processed, including gaussian filtering, histogram equalization；
Step 3:Binocular camera is demarcated, solves camera inside and outside parameter and camera distortion coefficient；
Step 4：Pixel coordinate is corrected by camera distortion coefficient, standard is done for image characteristic point detection and matching It is standby；
Step 5：Feature detection is carried out to image using ORB algorithms, cooperation FLANN carries out Feature Points Matching；
Step 6：It is constrained using epipolar geom etry and rejects Mismatching point, matching is optimized using RANSAC algorithms；
Step 7：The space of image characteristic point is solved by 3 d space coordinate calculation formula to the point coordinates after matching Coordinate obtains point cloud model, specific as follows：
According to three dimensional space coordinate formula (x1,y1,z1),
(u in formula1, v1) and (u2, v2) it is two point coordinates on binocular camera left images coordinate system, b is baseline length；
Step 8：Three-dimensional reconstruction is carried out to the spatial discrete points cloud of point cloud model using OpenGL.
In the step 3, including it is following step by step：
1) parametric array for needing to demarcate is established, including left and right according to the correspondence between feature using Zhang Zhengyou standardizations The relativeness matrix of camera intrinsic parameter matrix, distortion factor matrix and left and right camera；
2) homography matrix is solved：
Homography matrix：
More than there are four plane target drones, H can be solved：
λ[h1 h2 h3]=H=M1[r1 r2t]；
Wherein M be projection matrix, M1For camera Intrinsic Matrix, M2For Camera extrinsic matrix number；
It is constrained according to spin matrix, any two spin matrix vertically can obtain：
3) camera intrinsic parameter is solved：
Enable B=bij=M1 -TM1 -1That is symmetrical matrix；
Defined parameters variable b=[b11,b12,b13,b22,b23,b33]T, then have hi TBhj=Vij Tb；
Wherein
Have：6 internal references can be obtained as n >=3, b*=arg min can be solved | | Vb||；
It further can determine M1Each parameter；
4) according to M1The outer parameter of camera can further be obtained, be exactly M2R and T,
Coefficient lambda=1/ | | M1 -1h1| |=1/ | | M1 -1h2| |, r1=λ M1 -1h1, r2=λ M1 -1h2, r3=r1×r2,
T=λ M1 -1h3；
5) radial distortion parameter k1, k2 are further solved.
In the step 5, including it is following step by step：
1) ORB algorithms extract characteristic point using FAST algorithms, are sorted, taken N number of to characteristic point using Harris Corner Detections Preferable angle point is as characteristic point；
The method of ORB uses " intensity centroid " determines characteristic point direction, and the territory of characteristic point is seen Into a patch, the barycenter of this patch is asked for, by the barycenter of this patch and characteristic point line, the straight line and horizontal stroke is obtained The angle of reference axis is exactly the direction of this feature point；
The barycenter formula of the patch is as follows：
M in formulapqFor Gray Moment, I (x, y) is the gray value at (x, y) point of image, and p, q are the exponent number of Gray Moment；
2) barycenter is defined as：
M01,M10,M00The single order Gray Moment different for three；
3) direction of OC is asked for, x, the range of y is maintained within patch, using characteristic point as coordinate origin, obtains direction Angle is；
4) ORB algorithms solve rotational invariance and noise resisting ability, using the match point x in patch fieldsi,yiIt is if raw Into feature point description n test point be (xi,yi), the matrix for defining a 2*n is as follows：
Using spin matrix R, matched coordinate S after rotation is acquiredθ=RS,
5) characteristic point Detection and Extraction are carried out using ORB algorithms, while describes son generation characteristic point using improved Brief and retouch State symbol；
6) Feature Points Matching is carried out using FLANN algorithms.
In step 6, including it is following step by step：
1) by the video camera of distortion correction, on the polar curve of match point on the image plane；
2) P is calculatedl(xl,yl) and Pr(xr,yr) in | yl-yr| difference, and given threshold, threshold value is smaller then will to precision Ask higher；
3) match point except threshold range is rejected, correct match point is generally all near polar curve；
4) 4 groups of matching double points further screening and optimization are chosen from remaining match point, calculates homography matrix H, as Model；
5) projection error in data set with model is calculated, if less than threshold value, is incorporated as interior point set I；
6) optimal solution is determined whether according to interior number, if then point set I in update, while update iterations K；
7) if iterations are more than K, exit；Otherwise K+1 repeats the above steps；
Wherein p is confidence level, generally takes 0.995；W is the ratio of interior point；M is the required sample number 4 of computation model；
By the preliminary screening of epipolar-line constraint, the ratio w of interior point is relatively high, so as to reduce iterations, reduces operation When.
In step 8, including it is following step by step：
1) drafting of scene completion 3-D graphic is drawn using OpenGL functions in computer；
2) OpenGL functions is called to draw, is finally completed three reconstruct.
Technical scheme of the present invention is obtained using the binocular camera being mounted on unmanned machine head and transmission image, uses ORB image characteristic points detect and FLANN Feature Points Matchings, can accelerate system and perform speed, while have rotational invariance and resist The robustness of noise jamming greatlys improve the precision of real-time three-dimensional reconstruction.
Description of the drawings
Fig. 1 is schematic structural view of the invention.
Specific embodiment
The present invention is further illustrated below in conjunction with the accompanying drawings, but is not limitation of the invention.
Fig. 1 shows a kind of three-dimensional rebuilding method detected based on ORB features, double on unmanned machine head by being mounted on Mesh camera obtains realtime graphic and pretreatment, and video camera is demarcated, solves camera inside and outside parameter and distortion factor pair Pixel coordinate corrects, and image characteristic point is detected with the detection of ORB features, is matched using FLANN, the space of characteristic point is obtained Coordinate carries out three-dimensional reconstruction to spatial discrete points using OpenGL, includes the following steps：
Step 1:Binocular camera shooting system is mounted on unmanned machine head, establishes unmanned aerial vehicle base station system, is passed in real time Defeated image；
Step 2:The image of acquisition is pre-processed, including gaussian filtering, histogram equalization；
The noise that Gaussian filter is just being distributed very much inhibition is highly effective.Generally use 2-d gaussian filters function conduct Smoothing filter：
Step 3:Binocular camera is demarcated, solves camera inside and outside parameter and camera distortion coefficient；
Step 4：Pixel coordinate is corrected by camera distortion coefficient, standard is done for image characteristic point detection and matching It is standby；
Step 5：Feature detection is carried out to image using ORB algorithms, cooperation FLANN carries out Feature Points Matching；
Step 6：It is constrained using epipolar geom etry and rejects Mismatching point, matching is optimized using RANSAC algorithms；
Step 7：The space of image characteristic point is solved by 3 d space coordinate calculation formula to the point coordinates after matching Coordinate obtains point cloud model, specific as follows：
According to three dimensional space coordinate formula (x1,y1,z1),
(u in formula1, v1) and (u2, v2) it is two point coordinates on binocular camera left images coordinate system, b is baseline length；
Step 8：Three-dimensional reconstruction is carried out to the spatial discrete points cloud of point cloud model using OpenGL.
In the step 3, including it is following step by step：
1) parametric array for needing to demarcate is established, including left and right according to the correspondence between feature using Zhang Zhengyou standardizations The relativeness matrix of camera intrinsic parameter matrix, distortion factor matrix and left and right camera；
2) homography matrix is solved：
Homography matrix：
More than there are four plane target drones, H can be solved：
λ[h1 h2 h3]=H=M1[r1 r2t]；
Wherein M be projection matrix, M1For camera Intrinsic Matrix, M2For Camera extrinsic matrix number；
It is constrained according to spin matrix, any two spin matrix vertically can obtain：
3) camera intrinsic parameter is solved：
Enable B=bij=M1 -TM1 -1That is symmetrical matrix；
Defined parameters variable b=[b11,b12,b13,b22,b23,b33]T, then have hi TBhj=Vij Tb；
Wherein
Have：6 internal references can be obtained as n >=3, b*=arg min can be solved | | Vb||；
It further can determine M1Each parameter；
4) according to M1The outer parameter of camera can further be obtained, be exactly M2R and T,
Coefficient lambda=1/ | | M1 -1h1| |=1/ | | M1 -1h2| |, r1=λ M1 -1h1, r2=λ M1 -1h2, r3=r1×r2,
T=λ M1 -1h3；
5) radial distortion parameter k1, k2 are further solved.
In the step 5, including it is following step by step：
1) ORB algorithms extract characteristic point using FAST algorithms, are sorted, taken N number of to characteristic point using Harris Corner Detections Preferable angle point is as characteristic point；
The method of ORB uses " intensity centroid " determines characteristic point direction, and the territory of characteristic point is seen Into a patch, the barycenter of this patch is asked for, by the barycenter of this patch and characteristic point line, the straight line and horizontal stroke is obtained The angle of reference axis is exactly the direction of this feature point；
The barycenter formula of the patch is as follows：
M in formulapqFor Gray Moment, I (x, y) is the gray value at (x, y) point of image, and p, q are the exponent number of Gray Moment；
2) barycenter is defined as：
M01,M10,M00The single order Gray Moment different for three；
3) direction of OC is asked for, x, the range of y is maintained within patch, using characteristic point as coordinate origin, obtains direction Angle is；
4) ORB algorithms solve rotational invariance and noise resisting ability, using the match point x in patch fieldsi,yiIt is if raw Into feature point description n test point be (xi,yi), the matrix for defining a 2*n is as follows：
Using spin matrix R, matched coordinate S after rotation is acquiredθ=RS,
5) characteristic point Detection and Extraction are carried out using ORB algorithms, while describes son generation characteristic point using improved Brief and retouch State symbol；
6) Feature Points Matching is carried out using FLANN algorithms.
In step 6, including it is following step by step：
1) by the video camera of distortion correction, on the polar curve of match point on the image plane；
2) P is calculatedl(xl,yl) and Pr(xr,yr) in | yl-yr| difference, and given threshold, threshold value is smaller then will to precision Ask higher；
3) match point except threshold range is rejected, correct match point is generally all near polar curve；
4) 4 groups of matching double points further screening and optimization are chosen from remaining match point, calculates homography matrix H, as Model；
5) projection error in data set with model is calculated, if less than threshold value, is incorporated as interior point set I；
6) optimal solution is determined whether according to interior number, if then point set I in update, while update iterations K；
7) if iterations are more than K, exit；Otherwise K+1 repeats the above steps；
Wherein p is confidence level, generally takes 0.995；W is the ratio of interior point；M is the required sample number 4 of computation model；
By the preliminary screening of epipolar-line constraint, the ratio w of interior point is relatively high, so as to reduce iterations, reduces operation When.
In the step 8, including it is following step by step：
1) drafting of scene completion 3-D graphic is drawn using OpenGL functions in computer；
2) OpenGL functions is called to draw, is finally completed three reconstruct.
The advantageous effect of technical solution using the present invention can meet the binocular vision system three-dimensional reconstruction for carrying unmanned plane Requirement, it is ensured that accuracy and stability.
Embodiments of the present invention are made that with detailed description above in association with attached drawing, but the present invention be not limited to it is described Embodiment.To those skilled in the art, without departing from the principles and spirit of the present invention, these are implemented Mode carries out various change, modification, replacement and modification and still falls in protection scope of the present invention.

Claims (6)

1. a kind of three-dimensional rebuilding method based on the detection of ORB features, is obtained by the binocular camera being mounted on unmanned machine head Realtime graphic and pretreatment are taken, video camera is demarcated, solves camera inside and outside parameter and distortion factor to pixel coordinate school Just, it is detected with ORB features and image characteristic point is detected, matched using FLANN, the space coordinate of characteristic point is obtained, utilized OpenGL carries out three-dimensional reconstruction to spatial discrete points, which is characterized in that includes the following steps：
Step 1:Binocular camera shooting system is mounted on unmanned machine head, establishes unmanned aerial vehicle base station system, real-time Transmission figure Picture；
Step 2:The image of acquisition is pre-processed, including gaussian filtering, histogram equalization；
Step 3:Binocular camera is demarcated, solves camera inside and outside parameter and camera distortion coefficient；
Step 4：Pixel coordinate is corrected by camera distortion coefficient, is prepared for image characteristic point detection and matching；
Step 5：Feature detection is carried out to image using ORB algorithms, cooperation FLANN carries out Feature Points Matching；
Step 6：It is constrained using epipolar geom etry and rejects Mismatching point, matching is optimized using RANSAC algorithms；
Step 7：The space coordinate of image characteristic point is solved by 3 d space coordinate calculation formula to the point coordinates after matching Obtain point cloud model；
Step 8：Three-dimensional reconstruction is carried out to the spatial discrete points cloud of point cloud model using OpenGL.
2. the three-dimensional rebuilding method according to claim 1 based on the detection of ORB features, which is characterized in that the step 3 In, including it is following step by step：
1) parametric array for needing to demarcate is established, is imaged including left and right according to the correspondence between feature using Zhang Zhengyou standardizations The relativeness matrix of machine Intrinsic Matrix, distortion factor matrix and left and right camera；
2) homography matrix is solved：
Homography matrix：
More than there are four plane target drones, H can be solved：
λ[h1 h2 h3]=H=M1[r1 r2t]；
Wherein M be projection matrix, M1For camera Intrinsic Matrix, M2For Camera extrinsic matrix number；
It is constrained according to spin matrix, any two spin matrix vertically can obtain：
3) camera intrinsic parameter is solved：
Enable B=bij=M1 -TM1 -1That is symmetrical matrix；
Defined parameters variable b=[b11,b12,b13,b22,b23,b33]T, then have hi TBhj=Vij Tb；
Wherein
Have：6 internal references can be obtained as n >=3, b*=arg min can be solved | | Vb | |；
It further can determine M1Each parameter；
4) according to M1The outer parameter of camera can further be obtained, be exactly M2R and T,
Coefficient lambda=1/ | | M1 -1h1| |=1/ | | M1 -1h2| |, r1=λ M1 -1h1, r2=λ M1 -1h2, r3=r1×r2,
T=λ M1 -1h3；
5) radial distortion parameter k1, k2 are further solved.
3. the three-dimensional rebuilding method according to claim 1 based on the detection of ORB features, which is characterized in that the step 5 In, including it is following step by step：
1) ORB algorithms extract characteristic point using FAST algorithms, are sorted using Harris Corner Detections to characteristic point, take N number of preferable Angle point as characteristic point；
The method of ORB uses " intensity centroid " determines characteristic point direction, and the territory of characteristic point is regarded as one A patch asks for the barycenter of this patch, and by the barycenter of this patch and characteristic point line, the straight line and abscissa is obtained The angle of axis is exactly the direction of this feature point；
The barycenter formula of the patch is as follows：
M in formulapqFor Gray Moment, I (x, y) is the gray value at (x, y) point of image, and p, q are the exponent number of Gray Moment；
2) barycenter is defined as：
M01,M10,M00The single order Gray Moment different for three；
3) direction of OC is asked for, x, the range of y is maintained within patch, using characteristic point as coordinate origin, obtains deflection For；
4) ORB algorithms solve rotational invariance and noise resisting ability, using the match point x in patch fieldsi,yiIf generation N test point of feature point description is (xi,yi), the matrix for defining a 2*n is as follows：
Using spin matrix R, matched coordinate S after rotation is acquiredθ=RS,
5) characteristic point Detection and Extraction are carried out using ORB algorithms, while uses improved Brief description son generation feature point descriptions Symbol；
6) Feature Points Matching is carried out using FLANN algorithms.
4. the three-dimensional rebuilding method according to claim 1 based on the detection of ORB features, which is characterized in that the step 6 In, including it is following step by step：
1) by the video camera of distortion correction, on the polar curve of match point on the image plane；
2) P is calculatedl(xl,yl) and Pr(xr,yr) in | yl-yr| difference, and given threshold, threshold value is smaller, and required precision is got over It is high；
3) match point except threshold range is rejected, correct match point is generally all near polar curve；
4) 4 groups of matching double points further screening and optimization are chosen from remaining match point, homography matrix H is calculated, as model；
5) projection error in data set with model is calculated, if less than threshold value, is incorporated as interior point set I；
6) optimal solution is determined whether according to interior number, if then point set I in update, while update iterations K；
7) if iterations are more than K, exit；Otherwise K+1 repeats the above steps；
Wherein p is confidence level, generally takes 0.995；W is the ratio of interior point；M is the required sample number 4 of computation model；
By the preliminary screening of epipolar-line constraint, the ratio w of interior point is relatively high, so as to reduce iterations, when reducing operation.
5. the three-dimensional rebuilding method according to claim 1 based on the detection of ORB features, which is characterized in that the step 7 In, it is specific as follows：
According to three dimensional space coordinate formula (x1,y1,z1),
(u in formula1, v1) and (u2, v2) it is two point coordinates on binocular camera left images coordinate system, b is baseline length.
6. the three-dimensional rebuilding method according to claim 1 based on the detection of ORB features, which is characterized in that the step 8 In, including it is following step by step：
1) drafting of scene completion 3-D graphic is drawn using OpenGL functions in computer；
2) OpenGL functions is called to draw, is finally completed three reconstruct.
CN201711366005.6A 2017-12-18 2017-12-18 A kind of three-dimensional rebuilding method based on the detection of ORB features Pending CN108171787A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711366005.6A CN108171787A (en) 2017-12-18 2017-12-18 A kind of three-dimensional rebuilding method based on the detection of ORB features

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711366005.6A CN108171787A (en) 2017-12-18 2017-12-18 A kind of three-dimensional rebuilding method based on the detection of ORB features

Publications (1)

Publication Number Publication Date
CN108171787A true CN108171787A (en) 2018-06-15

Family

ID=62522364

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711366005.6A Pending CN108171787A (en) 2017-12-18 2017-12-18 A kind of three-dimensional rebuilding method based on the detection of ORB features

Country Status (1)

Country Link
CN (1) CN108171787A (en)

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109008909A (en) * 2018-07-13 2018-12-18 宜宾学院 A kind of low-power consumption capsule endoscope Image Acquisition and three-dimensional reconstruction system
CN109089100A (en) * 2018-08-13 2018-12-25 西安理工大学 A kind of synthetic method of binocular tri-dimensional video
CN109086795A (en) * 2018-06-27 2018-12-25 上海理工大学 A kind of accurate elimination method of image mismatch
CN109191509A (en) * 2018-07-25 2019-01-11 广东工业大学 A kind of virtual binocular three-dimensional reconstruction method based on structure light
CN109360269A (en) * 2018-09-30 2019-02-19 国网黑龙江省电力有限公司电力科学研究院 Three Dimensional Ground planar reconstruction method based on computer vision
CN109727287A (en) * 2018-12-27 2019-05-07 江南大学 A kind of improvement register method and its system suitable for augmented reality
CN109859314A (en) * 2019-03-12 2019-06-07 上海曼恒数字技术股份有限公司 Three-dimensional rebuilding method, device, electronic equipment and storage medium
CN109934911A (en) * 2019-03-15 2019-06-25 鲁东大学 Mobile terminal high-precision oblique photograph three-dimensional modeling method based on OpenGL
CN110009722A (en) * 2019-04-16 2019-07-12 成都四方伟业软件股份有限公司 Three-dimensional rebuilding method and device
CN110070568A (en) * 2019-04-28 2019-07-30 重庆学析优科技有限公司 A kind of picture antidote and system
WO2020024211A1 (en) * 2018-08-02 2020-02-06 深圳市道通智能航空技术有限公司 Unmanned aerial vehicle landing method and apparatus, and unmanned aerial vehicle
CN110909634A (en) * 2019-11-07 2020-03-24 深圳市凯迈生物识别技术有限公司 Visible light and double infrared combined rapid in vivo detection method
CN111080714A (en) * 2019-12-13 2020-04-28 太原理工大学 Parallel binocular camera calibration method based on three-dimensional reconstruction
CN111316325A (en) * 2019-03-08 2020-06-19 深圳市大疆创新科技有限公司 Shooting device parameter calibration method, equipment and storage medium
CN112002016A (en) * 2020-08-28 2020-11-27 中国科学院自动化研究所 Continuous curved surface reconstruction method, system and device based on binocular vision
CN112288852A (en) * 2020-10-28 2021-01-29 华润电力技术研究院有限公司 Coal yard three-dimensional reconstruction method and system and intelligent control method of thermal power generating unit
CN112347935A (en) * 2020-11-07 2021-02-09 南京天通新创科技有限公司 Binocular vision SLAM-based automatic driving vehicle positioning method and system
CN113228104A (en) * 2018-11-06 2021-08-06 菲力尔商业系统公司 Automatic co-registration of thermal and visible image pairs
WO2021195939A1 (en) * 2020-03-31 2021-10-07 深圳市大疆创新科技有限公司 Calibrating method for external parameters of binocular photographing device, movable platform and system
CN115359193A (en) * 2022-10-19 2022-11-18 南京航空航天大学 Rapid semi-dense three-dimensional reconstruction method based on binocular fisheye camera
CN117032276A (en) * 2023-07-04 2023-11-10 长沙理工大学 Bridge detection method and system based on binocular vision and inertial navigation fusion unmanned aerial vehicle

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103247075A (en) * 2013-05-13 2013-08-14 北京工业大学 Variational mechanism-based indoor scene three-dimensional reconstruction method
CN104732518A (en) * 2015-01-19 2015-06-24 北京工业大学 PTAM improvement method based on ground characteristics of intelligent robot

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103247075A (en) * 2013-05-13 2013-08-14 北京工业大学 Variational mechanism-based indoor scene three-dimensional reconstruction method
CN104732518A (en) * 2015-01-19 2015-06-24 北京工业大学 PTAM improvement method based on ground characteristics of intelligent robot

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
丁洁琼: "基于RGB-D的SLAM算法研究", 《万方数据库》 *
曹会敏: "用于目标识别的三维重建技术研究", 《万方数据库》 *

Cited By (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109086795A (en) * 2018-06-27 2018-12-25 上海理工大学 A kind of accurate elimination method of image mismatch
CN109008909A (en) * 2018-07-13 2018-12-18 宜宾学院 A kind of low-power consumption capsule endoscope Image Acquisition and three-dimensional reconstruction system
CN109008909B (en) * 2018-07-13 2024-01-26 宜宾学院 Low-power-consumption capsule endoscope image acquisition and three-dimensional reconstruction system
CN109191509A (en) * 2018-07-25 2019-01-11 广东工业大学 A kind of virtual binocular three-dimensional reconstruction method based on structure light
WO2020024211A1 (en) * 2018-08-02 2020-02-06 深圳市道通智能航空技术有限公司 Unmanned aerial vehicle landing method and apparatus, and unmanned aerial vehicle
CN109089100A (en) * 2018-08-13 2018-12-25 西安理工大学 A kind of synthetic method of binocular tri-dimensional video
CN109089100B (en) * 2018-08-13 2020-10-23 西安理工大学 Method for synthesizing binocular stereo video
CN109360269A (en) * 2018-09-30 2019-02-19 国网黑龙江省电力有限公司电力科学研究院 Three Dimensional Ground planar reconstruction method based on computer vision
US11869204B2 (en) 2018-11-06 2024-01-09 Teledyne Flir Commercial Systems, Inc. Automatic co-registration of thermal and visible image pairs
CN113228104A (en) * 2018-11-06 2021-08-06 菲力尔商业系统公司 Automatic co-registration of thermal and visible image pairs
CN109727287A (en) * 2018-12-27 2019-05-07 江南大学 A kind of improvement register method and its system suitable for augmented reality
CN109727287B (en) * 2018-12-27 2023-08-08 江南大学 Improved registration method and system suitable for augmented reality
CN111316325A (en) * 2019-03-08 2020-06-19 深圳市大疆创新科技有限公司 Shooting device parameter calibration method, equipment and storage medium
WO2020181409A1 (en) * 2019-03-08 2020-09-17 深圳市大疆创新科技有限公司 Capture device parameter calibration method, apparatus, and storage medium
CN111316325B (en) * 2019-03-08 2021-07-30 深圳市大疆创新科技有限公司 Shooting device parameter calibration method, equipment and storage medium
CN109859314A (en) * 2019-03-12 2019-06-07 上海曼恒数字技术股份有限公司 Three-dimensional rebuilding method, device, electronic equipment and storage medium
CN109934911A (en) * 2019-03-15 2019-06-25 鲁东大学 Mobile terminal high-precision oblique photograph three-dimensional modeling method based on OpenGL
CN110009722A (en) * 2019-04-16 2019-07-12 成都四方伟业软件股份有限公司 Three-dimensional rebuilding method and device
CN110070568A (en) * 2019-04-28 2019-07-30 重庆学析优科技有限公司 A kind of picture antidote and system
CN110909634A (en) * 2019-11-07 2020-03-24 深圳市凯迈生物识别技术有限公司 Visible light and double infrared combined rapid in vivo detection method
CN111080714A (en) * 2019-12-13 2020-04-28 太原理工大学 Parallel binocular camera calibration method based on three-dimensional reconstruction
WO2021195939A1 (en) * 2020-03-31 2021-10-07 深圳市大疆创新科技有限公司 Calibrating method for external parameters of binocular photographing device, movable platform and system
CN112002016A (en) * 2020-08-28 2020-11-27 中国科学院自动化研究所 Continuous curved surface reconstruction method, system and device based on binocular vision
CN112002016B (en) * 2020-08-28 2024-01-26 中国科学院自动化研究所 Continuous curved surface reconstruction method, system and device based on binocular vision
CN112288852A (en) * 2020-10-28 2021-01-29 华润电力技术研究院有限公司 Coal yard three-dimensional reconstruction method and system and intelligent control method of thermal power generating unit
CN112347935A (en) * 2020-11-07 2021-02-09 南京天通新创科技有限公司 Binocular vision SLAM-based automatic driving vehicle positioning method and system
CN112347935B (en) * 2020-11-07 2021-11-02 的卢技术有限公司 Binocular vision SLAM-based automatic driving vehicle positioning method and system
CN115359193A (en) * 2022-10-19 2022-11-18 南京航空航天大学 Rapid semi-dense three-dimensional reconstruction method based on binocular fisheye camera
CN115359193B (en) * 2022-10-19 2023-01-31 南京航空航天大学 Rapid semi-dense three-dimensional reconstruction method based on binocular fisheye camera
CN117032276A (en) * 2023-07-04 2023-11-10 长沙理工大学 Bridge detection method and system based on binocular vision and inertial navigation fusion unmanned aerial vehicle
CN117032276B (en) * 2023-07-04 2024-06-25 长沙理工大学 Bridge detection method and system based on binocular vision and inertial navigation fusion unmanned aerial vehicle

Similar Documents

Publication Publication Date Title
CN108171787A (en) A kind of three-dimensional rebuilding method based on the detection of ORB features
CN105956539B (en) A kind of Human Height measurement method of application background modeling and Binocular Vision Principle
CN110569704B (en) Multi-strategy self-adaptive lane line detection method based on stereoscopic vision
CN107292921B (en) Rapid three-dimensional reconstruction method based on kinect camera
CN106651942B (en) Three-dimensional rotating detection and rotary shaft localization method based on characteristic point
CN110992263B (en) Image stitching method and system
CN107679537B (en) A kind of texture-free spatial target posture algorithm for estimating based on profile point ORB characteristic matching
CN110956661B (en) Method for calculating dynamic pose of visible light and infrared camera based on bidirectional homography matrix
CN109685913B (en) Augmented reality implementation method based on computer vision positioning
CN108447022B (en) Moving target joining method based on single fixing camera image sequence
CN104144282B (en) A kind of fast digital digital image stabilization method suitable for robot for space vision system
CN110517349A (en) A kind of 3D vehicle target detection method based on monocular vision and geometrical constraint
CN111027415B (en) Vehicle detection method based on polarization image
CN106534833B (en) A kind of double vision point three-dimensional video-frequency antihunt means of joint space time shaft
CN107016646A (en) One kind approaches projective transformation image split-joint method based on improved
CN114255286A (en) Target size measuring method based on multi-view binocular vision perception
CN111553939B (en) Image registration algorithm of multi-view camera
CN107843251A (en) The position and orientation estimation method of mobile robot
CN107798702A (en) A kind of realtime graphic stacking method and device for augmented reality
CN108491810A (en) Vehicle limit for height method and system based on background modeling and binocular vision
CN109087323A (en) A kind of image three-dimensional vehicle Attitude estimation method based on fine CAD model
CN113313659B (en) High-precision image stitching method under multi-machine cooperative constraint
CN106920276A (en) A kind of three-dimensional rebuilding method and system
CN106971408A (en) A kind of camera marking method based on space-time conversion thought
CN115239820A (en) Split type flying vehicle aerial view real-time splicing and parking space detection method

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20180615