CN112884838A - Robot autonomous positioning method - Google Patents
Robot autonomous positioning method Download PDFInfo
- Publication number
- CN112884838A CN112884838A CN202110282300.3A CN202110282300A CN112884838A CN 112884838 A CN112884838 A CN 112884838A CN 202110282300 A CN202110282300 A CN 202110282300A CN 112884838 A CN112884838 A CN 112884838A
- Authority
- CN
- China
- Prior art keywords
- camera
- point
- image
- pose
- robot
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
- G06T7/74—Determining position or orientation of objects or cameras using feature-based methods involving reference images or patches
-
- 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
- G06T7/337—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving reference images or patches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/77—Determining position or orientation of objects or cameras using statistical methods
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20076—Probabilistic image processing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Manipulator (AREA)
Abstract
The invention discloses an autonomous robot positioning method, which comprises the following steps: 1) the robot collects the current environment image through a camera; 2) converting a current frame image collected by a camera and a selected reference image as a positioning reference into an HSI color space; 3) extraction of a point P in real environment space in a reference imagejProjected points in a reference image4) Calculating a point PjProjection point of current frame image5) Calculating projected pointsAnd projection pointHas a projection error of rj(ii) a 6) And continuously iterating and solving by minimizing the objective function E (xi) to obtain the optimal solution of the camera pose. According to the robot autonomous positioning method, the accuracy of pose estimation of the robot can be improved by utilizing more image information and constraint relation through the image registration result with higher accuracy, namely the accuracy of robot autonomous positioning is improved.
Description
Technical Field
The invention relates to the technical field of computer vision and robots, in particular to a robot autonomous positioning method.
Background
In the existing robot autonomous positioning method, a visual SLAM direct method is adopted, and a luminosity invariance assumed condition is used, so that after an input picture is directly converted into a gray image, the motion of a camera and the projection of a point are simultaneously estimated according to the pixel gray information of the image. However, according to the visual characteristics of human eyes, the human eyes are more sensitive to colors than gray scales; and the gray scale assumption condition can be influenced by the automatic exposure of the camera and the specular reflection of the object surface in the actual camera imaging system. Therefore, using the gradation information alone may cause image alignment failure.
In brief, image alignment, which aims to find the optimal image transformation to establish the spatial correspondence between different images, is widely applied in the field of computer vision. In addition to the visual SLAM field, many fields such as image depth estimation three-dimensional reconstruction, visual tracking and the like are established on the basis of image alignment, and the corresponding relation between images needs to be estimated by utilizing the image alignment. If the image alignment fails or is inaccurate, the accuracy of camera pose estimation, pixel point depth estimation value and the like is directly influenced.
Therefore, the method and the device aim at positioning requirements in the fields of robots and the like in practice, a more accurate and robust image alignment result is beneficial to improving the pose estimation precision of the robots, and play an important role in solving the positioning problem of the robots.
Disclosure of Invention
In view of the above, in order to solve the existing problems described above, an object of the present invention is to provide an autonomous positioning method for a robot, so as to solve the technical problem that the positioning accuracy of the robot is affected due to inaccurate image alignment in the existing autonomous positioning technology for the robot.
The invention discloses an autonomous robot positioning method, which comprises the following steps:
1) the robot collects the current environment image through a camera;
2) converting a current frame image collected by a camera and a selected reference image serving as a positioning reference into an HSI color space to obtain three components of H, S and I;
3) extraction of a point P in real environment space in a reference imagejProjected points in a reference image
In the above formula (1), K is camera reference, Z1Is a point PjDepth coordinate value, P, in reference frame camera coordinate systemj=[Xj,Yj,Zj]∈R3(ii) a C is a conversion matrix from homogeneous coordinates to non-homogeneous coordinates,
In the above formula (2), Z2Is a point PjDepth coordinate values under a current frame camera coordinate system, R is a pose rotation amount estimated value of a current frame image relative to a reference image, t is a pose translation amount estimated value of the current frame image relative to the reference image, and xi is a lie algebra corresponding to a camera pose (R, t), so that the lie algebra xi is used for representing camera pose amount;
wherein ξ ^ is an antisymmetric matrix of ξ;
5) calculating projected pointsAnd projection pointProjection error of rjThe projection error comprises a projection luminosity error eIjAnd projection color error eHj;
6) Suppose the same point PjThe gray values and color values under the current frame image and the reference image are constant,by minimizing the objective function E (xi), continuously iterating and solving to obtain the optimal solution xi of the camera pose*:
Further, the iterative solution in step 6) includes the following steps:
Step2 based on the derivative matrix JjCalculating the step length of descent, i.e. attitude increment delta, by Gauss-Newton methodξ:
Step3, updating the camera pose quantity ξ:
ξ=ξ0+δξ
wherein ξ0Representing the last iteration-derived camera pose quantity, δξRepresenting the increment of the camera pose, and xi representing the current camera pose quantity;
and (5) Setp4, repeatedly executing the Step1-3 until a convergence condition is met, ending the circulation to obtain the optimal pose solution ξ of the camera*=ξ。
The invention has the beneficial effects that:
the robot autonomous positioning method disclosed by the invention, 1) the pixel point errors of the luminosity component and the color component are combined, and as more image information is utilized, the image matching constraint relation is enhanced, and the matching precision of the multi-view image in the illumination change environment is improved; 2) the weighted value is added in the overall optimization objective function, so that the sensitivity of pixel point errors to brightness changes can be relieved to a certain extent, and if the weighted value of the pixel points in the area with large image luminosity error changes is reduced, the image alignment accuracy under the conditions of camera exposure and the like can be improved.
Because the assumption that the gray value is unchanged in practice is difficult to satisfy, the robot autonomous positioning method can better estimate the camera motion and the projection of the point under the condition that the whole image is lightened or darkened according to more information of image pixel gray and color.
Therefore, the invention can improve the pose estimation precision of the robot by using more image information and constraint relation through the image registration result with higher precision, namely improve the autonomous positioning precision of the robot.
Drawings
Fig. 1 shows the relationship between three-dimensional space points and image projection points, and the five-pointed star in the figure represents a point in the three-dimensional space.
Detailed Description
The invention is further described below with reference to the figures and examples.
The robot autonomous positioning method in the embodiment comprises the following steps:
1) the robot collects the current environment image through the camera.
2) And converting the current frame image collected by the camera and the selected reference image as the positioning reference into an HSI color space to obtain three components of H, S and I.
3) Extraction of a point P in real environment space in a reference imagejProjection point p in reference image1 j:
Pixel pointHas a gray value ofThe color components are Representing projected pointsThe number of rows in the image array,representing projected pointsThe number of columns in the image array,i.e. projected pointsThe image coordinates of (a).
In the above formula (1), K is camera reference, Z1Is a point PjDepth coordinate value, P, in reference frame camera coordinate systemj=[Xj,Yj,Zj]∈R3(ii) a C is a conversion matrix from homogeneous coordinates to non-homogeneous coordinates,
Pixel pointHas a gray value ofThe color components are Is a projected pointThe number of rows in the image array,is a projected pointThe number of columns in the image array,i.e. projected pointsThe image coordinates of (a).
In the above formula (2), Z2Is a point PjDepth coordinate values under a current frame camera coordinate system, R is a pose rotation amount estimated value of a current frame image relative to a reference image, t is a pose translation amount estimated value of the current frame image relative to the reference image, and xi is a lie algebra corresponding to a camera pose (R, t), so that the lie algebra xi is used for representing camera pose amount;
wherein ξ^Is an antisymmetric matrix of ξ.
5) Calculating projected pointsAnd projection pointHas a projection error of rjThe projection error comprises a projection luminosity error eIjAnd projection color error eHj;
6) Suppose the same point PjThe gray value and the color value under the current frame image and the reference image are unchanged, and the camera pose quantity xi is obtained by minimizing an objective function E (xi):
wherein N represents a point PjThe number of (2);
optimal camera pose xi*Solving by maximizing the posterior probability function:
and (3) converting the posterior probability density function into the prior probability density function by using Bayes formula, and equivalently replacing the formula (5) by the following formula:
by making the partial derivative equal to 0, neglecting the motion information P (ξ) term, there are:
since equation (8) is actually a form of weighted least squares, the optimization function of equation (5) is equivalent to:
for each projection error amount rjIs given a weight coefficient wj(ii) a The optimization objective function in modification (4) is:
by minimizing an objective functionContinuously and iteratively solving to obtain an optimal solution of the camera pose, wherein the iteratively solving comprises the following steps:
Step2 based on the derivative matrix JjCalculating the step length of descent, i.e. attitude increment delta, by Gauss-Newton methodξ:
Step3, updating the camera pose quantity ξ:
ξ=ξ0+δξ
wherein ξ0Representing the last iteration-derived camera pose quantity, δξRepresenting the increment of the camera pose, and xi representing the current camera pose quantity;
and (5) Setp4, repeatedly executing the Step1-3 until a convergence condition is met, ending the circulation to obtain the optimal pose solution ξ of the camera*=ξ。
Because the camera is arranged on the robot, the current camera pose quantity, namely the current position of the robot, is obtained.
In the embodiment, the robot autonomous positioning method comprises the steps that 1) pixel point errors of luminosity components and color components are combined, and more image information is utilized, so that the image matching constraint relation is enhanced, and the matching precision of multi-view images in the illumination change environment is improved; 2) the weighted value is added in the overall optimization objective function, so that the sensitivity of pixel point errors to brightness changes can be relieved to a certain extent, and if the weighted value of the pixel points in the area with large image luminosity error changes is reduced, the image alignment accuracy under the conditions of camera exposure and the like can be improved.
Because the assumption of unchanged gray value is difficult to satisfy in practice, the robot autonomous positioning method in the embodiment can better estimate the camera motion and the projection of the point under the condition that the whole image is bright or dark according to more information of the gray value and the color of the image pixel.
Therefore, according to the autonomous robot positioning method, the accuracy of pose estimation of the robot can be improved by using more image information and constraint relation through the image registration result with higher accuracy, and the accuracy of autonomous robot positioning can also be improved.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.
Claims (2)
1. An autonomous positioning method of a robot, comprising the steps of:
1) the robot collects the current environment image through a camera;
2) converting a current frame image collected by a camera and a selected reference image serving as a positioning reference into an HSI color space to obtain three components of H, S and I;
3) extraction of a point P in real environment space in a reference imagejProjected points in a reference image
In the above formula (1), K is camera reference, Z1Is a point PjDepth coordinate value, P, in reference frame camera coordinate systemj=[Xj,Yj,Zj]∈R3(ii) a C is a conversion matrix from homogeneous coordinates to non-homogeneous coordinates,
In the above formula (2), Z2Is a point PjDepth coordinate values under a current frame camera coordinate system, R is a pose rotation amount estimated value of a current frame image relative to a reference image, t is a pose translation amount estimated value of the current frame image relative to the reference image, and xi is a lie algebra corresponding to a camera pose (R, t), so that the lie algebra xi is used for representing camera pose amount;
wherein ξ^Is an antisymmetric matrix of ξ;
5) calculating projected pointsAnd projection pointHas a projection error of rjThe projection error comprises a projection luminosity error eIjAnd projection color error eHj;
6) FalseLet a same point PjThe gray value and the color value under the current frame image and the reference image are unchanged, and the optimal solution xi of the camera pose is obtained by minimizing an objective function E (xi) and continuously iterating and solving*:
2. The robot autonomous positioning method according to claim 1, characterized in that: the iterative solution in the step 6) comprises the following steps:
Step2 based on the derivative matrix JjCalculating the step length of descent, i.e. attitude increment delta, by Gauss-Newton methodξ:
Step3, updating the camera pose quantity ξ:
ξ=ξ0+δξ
wherein ξ0Representing the last iteration-derived camera pose quantity, δξRepresenting the increment of the camera pose, and xi representing the current camera pose quantity;
and (5) Setp4, repeatedly executing the Step1-3 until a convergence condition is met, ending the circulation to obtain the optimal pose solution ξ of the camera*=ξ。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110282300.3A CN112884838B (en) | 2021-03-16 | 2021-03-16 | Robot autonomous positioning method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110282300.3A CN112884838B (en) | 2021-03-16 | 2021-03-16 | Robot autonomous positioning method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112884838A true CN112884838A (en) | 2021-06-01 |
CN112884838B CN112884838B (en) | 2022-11-15 |
Family
ID=76042651
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110282300.3A Active CN112884838B (en) | 2021-03-16 | 2021-03-16 | Robot autonomous positioning method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112884838B (en) |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101498581A (en) * | 2008-12-29 | 2009-08-05 | 北京航空航天大学 | Relative attitude determination method for spacecraft based on three coplanar points |
US20140294298A1 (en) * | 2013-03-26 | 2014-10-02 | Xerox Corporation | Method and system for inverse halftoning utilizing inverse projection of predicted errors |
CN105976343A (en) * | 2016-04-29 | 2016-09-28 | 广东小天才科技有限公司 | Image exposure correction method, device and intelligent equipment |
US20170294020A1 (en) * | 2016-04-07 | 2017-10-12 | The Boeing Company | Camera pose estimation |
CN107945184A (en) * | 2017-11-21 | 2018-04-20 | 安徽工业大学 | A kind of mount components detection method positioned based on color images and gradient projection |
CN109211241A (en) * | 2018-09-08 | 2019-01-15 | 天津大学 | The unmanned plane autonomic positioning method of view-based access control model SLAM |
CN109387204A (en) * | 2018-09-26 | 2019-02-26 | 东北大学 | The synchronous positioning of the mobile robot of dynamic environment and patterning process in faced chamber |
CN109405824A (en) * | 2018-09-05 | 2019-03-01 | 武汉契友科技股份有限公司 | A kind of multi-source perceptual positioning system suitable for intelligent network connection automobile |
CN109670411A (en) * | 2018-11-30 | 2019-04-23 | 武汉理工大学 | Based on the inland navigation craft point cloud data depth image processing method and system for generating confrontation network |
CN109801233A (en) * | 2018-12-27 | 2019-05-24 | 中国科学院西安光学精密机械研究所 | A kind of Enhancement Method suitable for true color remote sensing image |
CN110152293A (en) * | 2018-02-13 | 2019-08-23 | 腾讯科技(深圳)有限公司 | Manipulate the localization method of object and the localization method and device of device, game object |
CN110175523A (en) * | 2019-04-26 | 2019-08-27 | 南京华捷艾米软件科技有限公司 | A kind of self-movement robot animal identification and hide method and its storage medium |
CN111445526A (en) * | 2020-04-22 | 2020-07-24 | 清华大学 | Estimation method and estimation device for pose between image frames and storage medium |
CN112053383A (en) * | 2020-08-18 | 2020-12-08 | 东北大学 | Method and device for real-time positioning of robot |
CN113129451A (en) * | 2021-03-15 | 2021-07-16 | 北京航空航天大学 | Holographic three-dimensional image space quantitative projection method based on binocular vision positioning |
-
2021
- 2021-03-16 CN CN202110282300.3A patent/CN112884838B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101498581A (en) * | 2008-12-29 | 2009-08-05 | 北京航空航天大学 | Relative attitude determination method for spacecraft based on three coplanar points |
US20140294298A1 (en) * | 2013-03-26 | 2014-10-02 | Xerox Corporation | Method and system for inverse halftoning utilizing inverse projection of predicted errors |
US20170294020A1 (en) * | 2016-04-07 | 2017-10-12 | The Boeing Company | Camera pose estimation |
CN105976343A (en) * | 2016-04-29 | 2016-09-28 | 广东小天才科技有限公司 | Image exposure correction method, device and intelligent equipment |
CN107945184A (en) * | 2017-11-21 | 2018-04-20 | 安徽工业大学 | A kind of mount components detection method positioned based on color images and gradient projection |
CN110152293A (en) * | 2018-02-13 | 2019-08-23 | 腾讯科技(深圳)有限公司 | Manipulate the localization method of object and the localization method and device of device, game object |
CN109405824A (en) * | 2018-09-05 | 2019-03-01 | 武汉契友科技股份有限公司 | A kind of multi-source perceptual positioning system suitable for intelligent network connection automobile |
CN109211241A (en) * | 2018-09-08 | 2019-01-15 | 天津大学 | The unmanned plane autonomic positioning method of view-based access control model SLAM |
CN109387204A (en) * | 2018-09-26 | 2019-02-26 | 东北大学 | The synchronous positioning of the mobile robot of dynamic environment and patterning process in faced chamber |
CN109670411A (en) * | 2018-11-30 | 2019-04-23 | 武汉理工大学 | Based on the inland navigation craft point cloud data depth image processing method and system for generating confrontation network |
CN109801233A (en) * | 2018-12-27 | 2019-05-24 | 中国科学院西安光学精密机械研究所 | A kind of Enhancement Method suitable for true color remote sensing image |
CN110175523A (en) * | 2019-04-26 | 2019-08-27 | 南京华捷艾米软件科技有限公司 | A kind of self-movement robot animal identification and hide method and its storage medium |
CN111445526A (en) * | 2020-04-22 | 2020-07-24 | 清华大学 | Estimation method and estimation device for pose between image frames and storage medium |
CN112053383A (en) * | 2020-08-18 | 2020-12-08 | 东北大学 | Method and device for real-time positioning of robot |
CN113129451A (en) * | 2021-03-15 | 2021-07-16 | 北京航空航天大学 | Holographic three-dimensional image space quantitative projection method based on binocular vision positioning |
Non-Patent Citations (4)
Title |
---|
CHRISTIAN KERL等: "Robust odometry estimation for RGB-D cameras", 《2013 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION》 * |
吴丹: "基于深度视觉的室内机器人定位研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
薛方正等: "基于CPG的双足机器人多层步行控制器设计", 《控制与决策》 * |
高北斗: "面向工业机器人的双目视觉系统研究", 《中国优秀硕士学位论文全文数据库》 * |
Also Published As
Publication number | Publication date |
---|---|
CN112884838B (en) | 2022-11-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109344882B (en) | Convolutional neural network-based robot control target pose identification method | |
US11830216B2 (en) | Information processing apparatus, information processing method, and storage medium | |
CN109754432B (en) | Camera automatic calibration method and optical motion capture system | |
CN108362266B (en) | Auxiliary monocular vision measurement method and system based on EKF laser ranging | |
CN113330486A (en) | Depth estimation | |
CN110281240B (en) | Method and system for positioning and picking up glass of liquid crystal display screen and vision processing system | |
WO2021218542A1 (en) | Visual perception device based spatial calibration method and apparatus for robot body coordinate system, and storage medium | |
CN114283203B (en) | Calibration method and system of multi-camera system | |
CN107040695B (en) | satellite-borne video image stabilization method and system based on RPC positioning model | |
CN113763479B (en) | Calibration method of refraction and reflection panoramic camera and IMU sensor | |
CN109781068B (en) | Visual measurement system ground simulation evaluation system and method for space application | |
CN113296395A (en) | Robot hand-eye calibration method in specific plane | |
CN114714356A (en) | Method for accurately detecting calibration error of hand eye of industrial robot based on binocular vision | |
CN114494449A (en) | Visual calibration and alignment laminating method for special-shaped product lamination | |
CN112184809A (en) | Relative pose estimation method, device, electronic device and medium | |
CN113362377B (en) | VO weighted optimization method based on monocular camera | |
CN112884838B (en) | Robot autonomous positioning method | |
CN111275764A (en) | Depth camera visual mileage measurement method based on line segment shadow | |
KR101766823B1 (en) | Robust visual odometry system and method to irregular illumination changes | |
CN117152257A (en) | Method and device for multidimensional angle calculation of ground monitoring camera | |
JP2022011818A (en) | Information processing apparatus and control method thereof | |
CN113256711B (en) | Pose estimation method and system of monocular camera | |
CN114998444A (en) | Robot high-precision pose measurement system based on two-channel network | |
CN114241059A (en) | Synchronous calibration method for camera and light source in photometric stereo vision system | |
CN114998429A (en) | Robot positioning system, method, apparatus, and medium |
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 |