CN108534782B - Binocular vision system-based landmark map vehicle instant positioning method - Google Patents

Binocular vision system-based landmark map vehicle instant positioning method Download PDF

Info

Publication number
CN108534782B
CN108534782B CN201810337903.7A CN201810337903A CN108534782B CN 108534782 B CN108534782 B CN 108534782B CN 201810337903 A CN201810337903 A CN 201810337903A CN 108534782 B CN108534782 B CN 108534782B
Authority
CN
China
Prior art keywords
landmark
vehicle
dimensional
camera
image
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
CN201810337903.7A
Other languages
Chinese (zh)
Other versions
CN108534782A (en
Inventor
郑亚莉
程洪
邱少波
田环根
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.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
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 University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201810337903.7A priority Critical patent/CN108534782B/en
Publication of CN108534782A publication Critical patent/CN108534782A/en
Application granted granted Critical
Publication of CN108534782B publication Critical patent/CN108534782B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20—Instruments for performing navigational calculations
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/005—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 with correlation of navigation data from several sources, e.g. map or contour matching

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Navigation (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a binocular vision system-based landmark map vehicle instant positioning method, which comprises the steps of utilizing a binocular vision system to realize a vision odometer, calculating the relative positioning of a vehicle, and then utilizing a deep learning technology to detect landmarks in an image; and then, carrying out landmark retrieval in a landmark map database, carrying out accurate positioning on the vehicle under a terrestrial coordinate system by utilizing the retrieved landmarks and a nonlinear optimization algorithm, then converting a camera relative coordinate system and the terrestrial coordinate system, and correcting the vehicle positioning of the binocular vision odometer by utilizing the vehicle positioning calculated by the landmarks.

Description

Binocular vision system-based landmark map vehicle instant positioning method
Technical Field
The invention belongs to the technical field of vehicle positioning, and particularly relates to a binocular vision system-based landmark map vehicle instant positioning method.
Background
The landmark map is a high-precision map which removes redundant information from the map and stores the redundant information in a light weight mode, and can provide partial static target perception, or landmark perception, for the intelligent vehicle. At present, high-tech companies such as israel, japan, and usa are engaged in the research and development and collection of landmark maps, such as eye q map of Mobile, landmark image map of toyota, and lvl5 crowdsourcing pure visual map. How to realize the positioning of the intelligent vehicle by using a landmark map is continuously researched by researchers in various countries.
By searching for prior art patents, we find that the system and the method for generating the binocular camera-based high-precision visual positioning map (application number 2016100288342), "the visual positioning method and the visual positioning device (application number 201611069552.3)," the visual positioning method and the device (application number 201110371807.2), "the system and the method for fusing and positioning the intelligent vehicle with the panoramic map" (application number 201710150551. X). The binocular camera-based high-precision visual positioning map generation system and method record the feature points in the image in a visual mode, implant the image feature points into the map to form a feature point map, and are different from a landmark map with actual significance. The visual positioning method and the visual positioning device set a plurality of identification points on an identification object, set a plurality of identification points on the identification object, establish a linear equation set of three-dimensional coordinates of the identification points, and solve the three-dimensional coordinates of the identification points, thereby obtaining the attitude information of the identification object. However, this method has disadvantages in that: the representation points need to be arranged on the identification object points, and the representation points cannot be used for positioning the intelligent vehicle. The visual positioning method and the visual positioning device realize that a three-dimensional image of the target object is synthesized by adopting a two-dimensional image of a camera through an image acquisition module, and the three-dimensional coordinate and the posture of the target object in a space coordinate system are determined according to the three-dimensional image. The disadvantages of this method are: reconstructions that rely on monocular cameras lack true physical dimension information and cannot be used for intelligent vehicle positioning. The intelligent vehicle fusion positioning system and method introducing the panoramic map obtains the positioning result of the vehicle by matching the vehicle-mounted satellite positioning equipment, the inertial navigation equipment, the camera sensor and the panoramic map. The disadvantages of this method are: depending on a plurality of sensors, the cost is expensive to obtain the ideal precision.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a landmark map vehicle instant positioning method based on a binocular vision system.
In order to achieve the aim, the invention discloses a landmark map vehicle instant positioning method based on a binocular vision system, which is characterized by comprising the following steps:
(1) carrying out internal reference calibration on the binocular vision system, acquiring internal parameters of a left camera and a right camera in the binocular vision system, and constructing an internal reference matrix K;
(2) the binocular vision system acquires video stream, the position of the left camera for acquiring the first frame image is used as an origin and is marked as C0Let the pose at this time be
Figure GDA0003103340110000021
(3) Continuously performing camera visual tracking from the obtained video stream, reconstructing the instant relative pose of the camera
Figure GDA0003103340110000022
Obtaining the relative coordinate position C of the vehiclej；
(3.1) extracting the feature X from the left and right video frames at the current moment jR,XLThen, for the feature X of the left and right video framesR,XLPerforming fast matching to obtain the characteristic XkWhere k is 1,2, …, which indicates the number of matched features, and j is 1,2, …, which indicates the time when the video stream continues;
reconstruction of two-dimensional feature points X by means of a multi-vision geometric methodkCorresponding three-dimensional feature point PkCoordinates of (2)
Figure GDA0003103340110000023
(3.2) removing outer points in the reconstructed three-dimensional characteristic points, and then establishing a local sparse map M by using the residual three-dimensional characteristic points;
(3.3) by using the assumed camera motion model estimation, rapidly searching the possible range of the two-dimensional feature points in the left or right video frame at the j +1 th moment, and establishing the image feature pair at the j +1 th moment and the j +1 th moment
Figure GDA0003103340110000024
Wherein, XsIs XROr XL(x, y) represents the coordinate location of the two-dimensional feature point;
establishing corresponding relation of three-dimensional characteristics according to image characteristic pairs of previous and next frames
Figure GDA0003103340110000025
And
Figure GDA0003103340110000026
and calculate out
Figure GDA0003103340110000027
And
Figure GDA0003103340110000028
relative transformation relationship of
Figure GDA0003103340110000029
Then the singleton appears in
Figure GDA00031033401100000210
Adding the three-dimensional feature points into the local sparse map M, and optimizing the three-dimensional feature points in the local sparse map M by a nonlinear optimization SBA algorithm;
(3.4) establishing a local sparse map M and
Figure GDA00031033401100000211
corresponding three-dimensional feature points, and then calculating the relative pose of the left or right camera at the j +1 th moment
Figure GDA0003103340110000031
Obtaining the relative coordinate position of the vehicle;
(4) detecting whether the images shot by the left camera or the right camera contain landmarks or not, and positioning the absolute coordinates of the vehicle through the detected landmarks;
(4.1) detecting landmark blocks marked as B in each image shot by the left or right camera by using a deep learning method1,B2,…,BnN represents the number of landmark blocks in the image; recording the shooting time of the image of the detected landmark block as i, i equals to 1,2, …, i < j;
and then carrying out feature matching on the landmark blocks in the landmark map database to find out the coordinates of the landmark blocks under the global coordinate system
Figure GDA0003103340110000032
Namely longitude and latitude elevation;
(4.2) detecting the retrieved landmark block B1,B2,…,BnAnd calculating the coordinates of the two-dimensional image of the center of the landmark block in the video
Figure GDA0003103340110000033
Correcting the coordinates of the two-dimensional image by using the internal parameters of the camera; then, establishing two-dimensional image coordinates of the landmark block and three-dimensional coordinates of the landmark under the global coordinate system
Figure GDA0003103340110000034
Corresponding relation between them;
(4.3) establishing an optimized objective function F
Figure GDA0003103340110000035
Wherein R represents a rotation matrix, t represents a translation vector, d represents a Euclidean distance,
Figure GDA0003103340110000036
the coordinates of the two-dimensional image are represented,
Figure GDA0003103340110000037
representing two-dimensional projection vectors obtained by calculation under a homogeneous coordinate system;
obtaining the absolute coordinate position of the vehicle under the global coordinate system at the current moment i by optimally solving R and t
Figure GDA0003103340110000038
(5) Establishing the absolute coordinate position of the vehicle
Figure GDA0003103340110000039
Relative coordinate position of vehicle with corresponding time
Figure GDA00031033401100000310
Relative transformation relation of (T)i，
Figure GDA00031033401100000311
Representing a relative element corresponding to the ith moment when the landmark block is detected in the relative coordinate sequence; then use TiCorrecting the relative coordinate position of the vehicle to the absolute coordinate position of the vehicle under the global coordinate system, and correcting the positions of all the vehicles at the i-1 th moment and the i-th moment before the correction
Figure GDA00031033401100000312
To an absolute position
Figure GDA00031033401100000313
And finishing the instant absolute positioning of the vehicle.
The invention aims to realize the following steps:
the invention relates to a landmark map vehicle instant positioning method based on a binocular vision system, which is characterized in that a binocular vision system is utilized to realize a vision odometer, the relative positioning of a vehicle is calculated, and then a deep learning technology is utilized to detect landmarks in an image; and then, carrying out landmark retrieval in a landmark map database, carrying out accurate positioning on the vehicle under a terrestrial coordinate system by utilizing the retrieved landmarks and a nonlinear optimization algorithm, then converting a camera relative coordinate system and the terrestrial coordinate system, and correcting the vehicle positioning of the binocular vision odometer by utilizing the vehicle positioning calculated by the landmarks.
Meanwhile, the method for instantly positioning the landmark map vehicle based on the binocular vision system also has the following beneficial effects:
(1) the binocular vision system is used as the only sensor to realize the instant positioning function based on the landmark map vehicle, the advantages of low cost are achieved, the binocular vision system can be multiplexed with the vision perception module sensor of the intelligent vehicle, and the cost can be effectively reduced.
(2) Compared with the prior art, the invention has the advantages of improved performance, improved reliability, reduced cost, simplified process, energy saving, environmental protection, and the like.
Drawings
FIG. 1 is a binocular vision system based landmark map vehicle instant positioning system architecture diagram of the present invention;
FIG. 2 is a flow chart of the method for instantly positioning a landmark map vehicle based on a binocular vision system according to the present invention;
FIG. 3 is a schematic of a three-dimensional reconstruction;
fig. 4 is an effect diagram of the instant positioning of the vehicle by using the invention.
Detailed Description
The following description of the embodiments of the present invention is provided in order to better understand the present invention for those skilled in the art with reference to the accompanying drawings. It is to be expressly noted that in the following description, a detailed description of known functions and designs will be omitted when it may obscure the subject matter of the present invention.
Examples
Fig. 1 is a binocular vision system based landmark map vehicle instant positioning system architecture diagram of the present invention.
In this embodiment, as shown in fig. 1, the landmark map vehicle instant positioning system based on the binocular vision system of the present invention mainly includes:
binocular vision acquisition module: the binocular vision correction system is used for acquiring binocular vision signals, correcting images of a binocular camera and establishing a unified terrestrial coordinate system.
A landmark detection module: firstly, training a deep neural network model by using a deep learning method, then detecting landmarks in real time by using the deep neural network model, and returning coordinates of the landmarks in a two-dimensional image;
map database of landmarks: the method is used for downloading the landmark map, and the landmark map comprises image blocks of all landmarks and longitude and latitude altitude coordinates of the image blocks under the earth coordinate system.
The vehicle instant positioning module: the outputs of the binocular vision acquisition module, the landmark map database and the landmark detection module are used as the inputs, so that the function of vehicle instant positioning is realized.
In the following, we will describe in detail a method for instantly positioning a landmark map vehicle based on a binocular vision system in conjunction with fig. 2, which specifically includes the following steps:
s1, carrying out internal reference calibration on the binocular vision acquisition module to obtain internal parameters of a left camera and a right camera in the binocular vision system, wherein the internal parameters mainly comprise distortion parameters, focal length and center offset of the cameras, and then constructing an internal reference matrix K by utilizing the internal parameters.
S2, the binocular vision acquisition module acquires video stream, the position of the left camera for acquiring the first frame image is used as an origin and is marked as C0Let the pose at this time be
Figure GDA0003103340110000051
Mark C0And
Figure GDA0003103340110000052
for reference in subsequent coordinate rectification processes.
S3, continuously carrying out camera visual tracking from the obtained video stream, and reconstructing the instant relative posture of the camera
Figure GDA0003103340110000053
Obtaining the relative coordinate position C of the vehiclej；
S3.1, extracting the feature X from the left video frame and the right video frame by utilizing algorithms such as ORB, SURF, SIFT and the like at the current moment jR,XLAnd then the polar line search method is utilized to search the characteristics X of the left and right video framesR,XLPerforming fast matching to obtain the characteristic XkWhere k is 1,2, …, which indicates the number of matched features, and j is 1,2, …, which indicates the time when the video stream continues;
reconstruction of two-dimensional feature points X by means of a multi-vision geometric methodkCorresponding three-dimensional feature point PkCoordinates of (2)
Figure GDA0003103340110000054
S3.2, removing outer points in the reconstructed three-dimensional feature points, namely taking out the three-dimensional feature points with the reconstructed precision exceeding a threshold value, so that the accuracy of the reconstructed three-dimensional feature points is ensured, wherein the schematic diagram of the three-dimensional reconstructed three-dimensional feature points is shown in FIG. 3, and then establishing a local sparse map M by using the remaining three-dimensional feature points;
s3.3, by using the assumed camera motion model estimation, rapidly searching the possible range of the two-dimensional feature points in the left or right video frame at the j +1 th moment, and establishing the image feature pair at the j +1 th moment and the j +1 th moment
Figure GDA0003103340110000055
Wherein, XsIs XROr XL(x, y) represents the coordinate location of the two-dimensional feature point;
establishing three-dimension according to image characteristic pairs of front and back framesCorrespondence of features
Figure GDA0003103340110000056
And
Figure GDA0003103340110000057
and calculate out
Figure GDA0003103340110000058
And
Figure GDA0003103340110000059
relative transformation relationship of
Figure GDA00031033401100000510
Then the singleton appears in
Figure GDA00031033401100000511
Adding the three-dimensional feature points into the local sparse map M, and optimizing the three-dimensional feature points in the local sparse map M by a nonlinear optimization SBA algorithm;
s3.4, establishing a local sparse map M and
Figure GDA0003103340110000061
corresponding three-dimensional feature points, and then calculating the relative pose of the left or right camera at the j +1 th moment
Figure GDA0003103340110000062
Obtaining the relative coordinate position of the vehicle;
s4, detecting whether the images shot by the left camera or the right camera contain the landmark or not, and positioning the absolute coordinate of the vehicle through the detected landmark;
s4.1, detecting landmark blocks marked as B in each image shot by the left camera or the right camera by using a deep learning method1,B2,…,BnN represents the number of landmark blocks in the image; recording the shooting time of the image of the detected landmark block as i, i equals to 1,2, …, i < j;
then feature matching is carried out on the landmark blocks in the landmark map database, and a global coordinate system is foundCoordinates of underlying landmark blocks
Figure GDA0003103340110000063
Namely longitude and latitude elevation;
in this embodiment, in order to increase the speed of feature matching on landmark blocks, the present embodiment may perform landmark block feature matching in a limited range in a landmark map database, where the limited range may be provided by a non-high-precision satellite positioning system to assist in landmark search and matching;
s4.2, detecting the searched landmark block B1,B2,…,BnAnd calculating the coordinates of the two-dimensional image of the center of the landmark block in the video
Figure GDA0003103340110000064
Correcting the coordinates of the two-dimensional image by using the internal parameters of the camera; then, establishing two-dimensional image coordinates of the landmark block and three-dimensional coordinates of the landmark under the global coordinate system
Figure GDA0003103340110000065
Corresponding relation between them;
s4.3, establishing an optimized objective function F
Figure GDA0003103340110000066
Wherein R represents a rotation matrix, t represents a translation vector, d represents an Euclidean distance,
Figure GDA0003103340110000067
the coordinates of the two-dimensional image are represented,
Figure GDA0003103340110000068
representing two-dimensional projection vectors obtained by calculation under a homogeneous coordinate system;
obtaining the absolute coordinate position of the vehicle under the global coordinate system at the current moment i by optimally solving R and t
Figure GDA0003103340110000069
S5, establishing absolute coordinate position of vehicle
Figure GDA00031033401100000610
Relative coordinate position of vehicle with corresponding time
Figure GDA00031033401100000611
Relative transformation relation of (T)i，
Figure GDA00031033401100000612
Representing a relative element corresponding to the ith moment when the landmark block is detected in the relative coordinate sequence; then use TiCorrecting the relative coordinate position of the vehicle to the absolute coordinate position of the vehicle under the global coordinate system, and correcting the positions of all the vehicles at the i-1 th moment and the i-th moment before the correction
Figure GDA0003103340110000071
To an absolute position
Figure GDA0003103340110000072
And finishing the instant absolute positioning of the vehicle.
In this embodiment, i represents the time when the landmark is detected, and j represents the time when the video stream continues, so i<<j, we now use the absolute coordinate position of time i
Figure GDA0003103340110000073
Correcting the relative coordinate position of the corresponding i time in j
Figure GDA0003103340110000074
And using the absolute coordinate position of the time i
Figure GDA0003103340110000075
And correcting the relative coordinate position of the vehicle when the ground mark is not detected in the video stream before the time j. As shown in fig. 4, the experimental results of the vehicle instantaneous location system are shown, in which the left camera image is shown on the left and the instantaneous location results of the vehicle are shown on the right.
Although illustrative embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, and various changes may be made apparent to those skilled in the art as long as they are within the spirit and scope of the present invention as defined and defined by the appended claims, and all matters of the invention which utilize the inventive concepts are protected.

Claims (2)

1. A landmark map vehicle instant positioning method based on a binocular vision system is characterized by comprising the following steps:
(1) carrying out internal reference calibration on the binocular vision system, acquiring internal parameters of a left camera and a right camera in the binocular vision system, and constructing an internal reference matrix K;
(2) the binocular vision system obtains video stream, the position of the center of the camera for shooting the first image in the left camera is used as an origin, and the mark is C0Let the pose at this time be
Figure FDA0002952088180000011
(3) Continuously performing camera visual tracking from the obtained video stream, reconstructing the instant relative pose of the camera
Figure FDA0002952088180000012
Obtaining the relative coordinate position C of the vehiclej；
(3.1) extracting the feature X from the left and right video frames at the current moment jR,XLThen, for the feature X of the left and right video framesR,XLPerforming fast matching to obtain the characteristic XkWherein, k is 1,2, …, which represents the number of matched features;
reconstruction of two-dimensional feature points X by means of a multi-vision geometric methodkCorresponding three-dimensional feature point PkCoordinates of (2)
Figure FDA0002952088180000013
(3.2) removing outer points in the reconstructed three-dimensional characteristic points, and then establishing a local sparse map M by using the residual three-dimensional characteristic points;
(3.3) by using the assumed camera motion model estimation, rapidly searching the possible range of the two-dimensional feature points in the left or right video frame at the j +1 th moment, and establishing the image feature pair at the j +1 th moment and the j +1 th moment
Figure FDA0002952088180000014
Wherein, XsIs XROr XL(x, y) represents the coordinate position of the two-dimensional feature point;
establishing corresponding relation of three-dimensional characteristics according to image characteristic pairs of previous and next frames
Figure FDA0002952088180000015
And
Figure FDA0002952088180000016
and calculate out
Figure FDA0002952088180000017
And
Figure FDA0002952088180000018
relative transformation relationship of
Figure FDA0002952088180000019
Then the singleton appears in
Figure FDA00029520881800000110
Adding the three-dimensional feature points into the local sparse map M, and optimizing the three-dimensional feature points in the local sparse map M by a nonlinear optimization SBA algorithm;
(3.4) establishing a local sparse map M and
Figure FDA00029520881800000111
corresponding three-dimensional feature points, and then calculating the relative pose of the left or right camera at the j +1 th moment
Figure FDA00029520881800000112
Obtaining the relative coordinate position of the vehicle;
(4) detecting whether the images shot by the left camera or the right camera contain landmarks or not, and positioning the absolute coordinates of the vehicle through the detected landmarks;
(4.1) detecting landmark blocks marked as B in each image shot by the left or right camera by using a deep learning method1,B2,…,BnN represents the number of landmark blocks in the image; recording the shooting time of the image of the detected landmark block as i, i equals to 1,2, …, i < j;
and then carrying out feature matching on the landmark blocks in the landmark map database to find out the coordinates of the landmark blocks under the global coordinate system
Figure FDA0002952088180000021
Namely longitude and latitude elevation;
(4.2) detecting the retrieved landmark block B1,B2,…,BnAnd calculating the coordinates of the two-dimensional image of the center of the landmark block in the video
Figure FDA0002952088180000022
Correcting the coordinates of the two-dimensional image by using the internal parameters of the camera; then, establishing two-dimensional image coordinates of the landmark block and three-dimensional coordinates of the landmark under the global coordinate system
Figure FDA0002952088180000023
Corresponding relation between them;
(4.3) establishing an optimization objective function F of the error between the two-dimensional image coordinate of the landmark block and the three-dimensional coordinate of the landmark in the global coordinate system;
Figure FDA0002952088180000024
wherein R represents a rotation matrix, t represents a translation vector, d represents a Euclidean distance,
Figure FDA0002952088180000025
the coordinates of the two-dimensional image are represented,
Figure FDA0002952088180000026
representing two-dimensional projection vectors obtained by calculation under a homogeneous coordinate system;
obtaining the absolute coordinate position of the vehicle under the global coordinate system at the current moment i by optimally solving R and t
Figure FDA0002952088180000027
(5) Establishing the absolute coordinate position of the vehicle
Figure FDA0002952088180000028
Relative coordinate position of vehicle with corresponding time
Figure FDA0002952088180000029
Relative transformation relation of (T)i，
Figure FDA00029520881800000210
Representing a relative element corresponding to the ith moment when the landmark block is detected in the relative coordinate sequence; then use TiCorrecting the relative coordinate position of the vehicle to the absolute coordinate position of the vehicle under the global coordinate system, and correcting the positions of all the vehicles at the i-1 th moment and the i-th moment before the correction
Figure FDA00029520881800000211
To an absolute position
Figure FDA00029520881800000212
And finishing the instant absolute positioning of the vehicle.
2. The binocular vision system-based landmark map vehicle instant positioning method of claim 1, wherein the left and right camera intrinsic parameters comprise distortion parameters, focal length and center offset of the cameras.
CN201810337903.7A 2018-04-16 2018-04-16 Binocular vision system-based landmark map vehicle instant positioning method Active CN108534782B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810337903.7A CN108534782B (en) 2018-04-16 2018-04-16 Binocular vision system-based landmark map vehicle instant positioning method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810337903.7A CN108534782B (en) 2018-04-16 2018-04-16 Binocular vision system-based landmark map vehicle instant positioning method

Publications (2)

Publication Number Publication Date
CN108534782A CN108534782A (en) 2018-09-14
CN108534782B true CN108534782B (en) 2021-08-17

Family

ID=63481245

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810337903.7A Active CN108534782B (en) 2018-04-16 2018-04-16 Binocular vision system-based landmark map vehicle instant positioning method

Country Status (1)

Country Link
CN (1) CN108534782B (en)

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110706280A (en) * 2018-09-28 2020-01-17 成都家有为力机器人技术有限公司 Lightweight semantic driven sparse reconstruction method based on 2D-SLAM
CN109584299B (en) * 2018-11-13 2021-01-05 深圳前海达闼云端智能科技有限公司 Positioning method, positioning device, terminal and storage medium
CN109583409A (en) * 2018-12-07 2019-04-05 电子科技大学 A kind of intelligent vehicle localization method and system towards cognitive map
TWM584008U (en) * 2019-01-31 2019-09-21 許斐凱 Trachea model reconstruction system utilizing computer vision and deep learning technology
CN109752008B (en) * 2019-03-05 2021-04-13 长安大学 Intelligent vehicle multi-mode cooperative positioning system and method and intelligent vehicle
CN110514212A (en) * 2019-07-26 2019-11-29 电子科技大学 A kind of intelligent vehicle map terrestrial reference localization method merging monocular vision and difference GNSS
CN112307810B (en) * 2019-07-26 2023-08-04 北京魔门塔科技有限公司 Visual positioning effect self-checking method and vehicle-mounted terminal
CN111539973B (en) * 2020-04-28 2021-10-01 北京百度网讯科技有限公司 Method and device for detecting pose of vehicle
CN112132894B (en) * 2020-09-08 2022-09-20 大连理工大学 Mechanical arm real-time tracking method based on binocular vision guidance
CN112489080A (en) * 2020-11-27 2021-03-12 的卢技术有限公司 Binocular vision SLAM-based vehicle positioning and vehicle 3D detection method
CN112598705B (en) * 2020-12-17 2024-05-03 太原理工大学 Binocular vision-based vehicle body posture detection method
CN113390435B (en) * 2021-05-13 2022-08-26 中铁二院工程集团有限责任公司 High-speed railway multi-element auxiliary positioning system
CN114998849B (en) * 2022-05-27 2024-04-16 电子科技大学 Traffic flow element sensing and positioning method based on road-side monocular camera and application thereof
CN116528062B (en) * 2023-07-05 2023-09-15 合肥中科类脑智能技术有限公司 Multi-target tracking method

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8818133B2 (en) * 2012-07-11 2014-08-26 Raytheon Company Point cloud construction with unposed camera
CN103106688B (en) * 2013-02-20 2016-04-27 北京工业大学 Based on the indoor method for reconstructing three-dimensional scene of double-deck method for registering
CN105469405B (en) * 2015-11-26 2018-08-03 清华大学 Positioning and map constructing method while view-based access control model ranging
CN106228538B (en) * 2016-07-12 2018-12-11 哈尔滨工业大学 Binocular vision indoor orientation method based on logo
CN106709950B (en) * 2016-11-28 2020-09-22 西安工程大学 Binocular vision-based inspection robot obstacle crossing wire positioning method
CN106910210B (en) * 2017-03-03 2018-09-11 百度在线网络技术（北京）有限公司 Method and apparatus for generating image information

Also Published As

Publication number Publication date
CN108534782A (en) 2018-09-14

Similar Documents

Publication Publication Date Title
CN108534782B (en) Binocular vision system-based landmark map vehicle instant positioning method
CN108801274B (en) Landmark map generation method integrating binocular vision and differential satellite positioning
CN110070615B (en) Multi-camera cooperation-based panoramic vision SLAM method
CN110009681B (en) IMU (inertial measurement unit) assistance-based monocular vision odometer pose processing method
CN111830953B (en) Vehicle self-positioning method, device and system
CN106679648B (en) Visual inertia combination SLAM method based on genetic algorithm
CN110319772B (en) Visual large-span distance measurement method based on unmanned aerial vehicle
US10909395B2 (en) Object detection apparatus
CN109579825B (en) Robot positioning system and method based on binocular vision and convolutional neural network
CN109596121B (en) Automatic target detection and space positioning method for mobile station
CN107560603B (en) Unmanned aerial vehicle oblique photography measurement system and measurement method
CN113850126A (en) Target detection and three-dimensional positioning method and system based on unmanned aerial vehicle
CN108519102B (en) Binocular vision mileage calculation method based on secondary projection
US20070115352A1 (en) System and method for multi-camera visual odometry
CN104281148A (en) Mobile robot autonomous navigation method based on binocular stereoscopic vision
CN113658337B (en) Multi-mode odometer method based on rut lines
CN111815765B (en) Heterogeneous data fusion-based image three-dimensional reconstruction method
CN111091088B (en) Video satellite information supported marine target real-time detection positioning system and method
CN117253029B (en) Image matching positioning method based on deep learning and computer equipment
CN113240813B (en) Three-dimensional point cloud information determining method and device
CN114693754B (en) Unmanned aerial vehicle autonomous positioning method and system based on monocular vision inertial navigation fusion
CN111524174A (en) Binocular vision three-dimensional construction method for moving target of moving platform
CN112767546A (en) Binocular image-based visual map generation method for mobile robot
CN116385504A (en) Inspection and ranging method based on unmanned aerial vehicle acquisition point cloud and image registration
CN113345032B (en) Initialization map building method and system based on wide-angle camera large distortion map

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