CN114037800A - Construction system, method and device of octree map and electronic equipment - Google Patents
Construction system, method and device of octree map and electronic equipment Download PDFInfo
- Publication number
- CN114037800A CN114037800A CN202111354683.7A CN202111354683A CN114037800A CN 114037800 A CN114037800 A CN 114037800A CN 202111354683 A CN202111354683 A CN 202111354683A CN 114037800 A CN114037800 A CN 114037800A
- Authority
- CN
- China
- Prior art keywords
- information
- point cloud
- pose
- map
- laser radar
- 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
Links
Images
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
- G06T17/005—Tree description, e.g. octree, quadtree
-
- 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
- G06T17/05—Geographic models
-
- 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
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Remote Sensing (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Traffic Control Systems (AREA)
- Navigation (AREA)
Abstract
The invention provides a construction system, a construction method and a construction device of an octree map and electronic equipment. The method comprises the following steps: the system comprises a laser radar sensor, a data processing unit and a data processing unit, wherein the laser radar sensor is used for acquiring point cloud data of the surrounding environment; the inertial navigation unit is used for acquiring acceleration information and angular velocity information of the equipment; the data processing control unit is connected with the laser radar sensor and the inertial navigation unit, performs pre-integration on the acceleration information and the angular velocity information to obtain current pose information of the laser radar sensor, and performs distortion correction on point cloud data through the current pose information; detecting characteristic points of data in the point cloud by a curvature method, utilizing the characteristic points to perform scene correlation calculation on pose transformation information of the laser radar sensor and position information of the current point cloud in space, and creating and updating an octree map according to the pose transformation information and the position information. The construction system of the octree map solves the problem that the precision of estimating the motion of the point cloud by using the octree map is poor in the prior art.
Description
Technical Field
The invention relates to the technical field of scene association and construction of three-dimensional unknown space, in particular to a construction system, a construction method, a construction device and electronic equipment of an octree map.
Background
Simultaneous localization and mapping (SLAM) are hot and difficult points of robot research, and can be mainly classified into 2D SLAM and 3D SLAM according to the type of laser ranging sensor. For 2D SLAM, a grid map is a map form commonly used at present, the grid map is more beneficial to positioning and navigation of a robot, but the positioning and navigation of the 2D SLAM are usually dependent on a two-dimensional plane, and no better solution is provided for height change of a sensor, so that the 3D SLAM is more suitable for practical use.
For 3D SLAM, currently, main research focuses on estimating robot pose, and the frequently adopted map form is often a feature point map, for example: although mainstream algorithms such as LOAM, Lego-LOAM and LIO-SAM are simpler and more convenient for the pose change calculation of the robot, the defects of the point feature point cloud map are relatively obvious. Firstly, the scale of the characteristic point cloud map is relatively large, and a lot of unnecessary details are provided, so that the waste of storage space is caused; secondly, because the characteristic point cloud map is formed by overlapping points, the point cloud map is difficult to process for moving objects.
The octree map (OctMap) is a compressed map, the information of whether the map is occupied or not is stored through the data structure of the octree, the unoccupied points of the root nodes are not expanded, so the space storage efficiency is higher, the octree map is relatively simple to represent, store and update compared with a characteristic point map, and the map information can be updated in time through the update of the occupation probability of the map for moving objects. When the robot navigates, the sensing of the obstacle is more facilitated. For the constructed octree map, the height of the robot sensor is not required to be limited, and the method can be used for guiding the unmanned aerial vehicle. The existing octree map creation mainly converts the created feature point map into the whole map, and although the octree map creation can be used for navigation and display of environmental structures, the advantages of storage efficiency and map updating are lost. In addition, compared with the method for estimating the pose change of the sensor by using the characteristic points of the point cloud, the method for estimating the motion of the point cloud by using the octree map has poor precision.
Disclosure of Invention
The invention aims to provide a construction system, a construction method, a construction device and electronic equipment of an octree map.
In order to achieve the above purpose, the invention provides the following technical scheme:
a construction system of an octree map, comprising:
the system comprises a laser radar sensor, a data processing unit and a data processing unit, wherein the laser radar sensor is used for acquiring point cloud data of the surrounding environment;
the inertial navigation unit is used for acquiring acceleration information and angular velocity information of the equipment;
the data processing control unit is connected with the laser radar sensor and the inertial navigation unit, performs pre-integration on acceleration information and angular velocity information to obtain current pose information of the laser radar sensor, and performs distortion correction on point cloud data through the current pose information; detecting feature points of data in the point cloud by a curvature method, utilizing the feature points to perform scene correlation calculation on pose transformation information of the laser radar sensor and position information of the current point cloud in space, and creating and updating an octree map according to the pose transformation information and the position information.
On the basis of the technical scheme, the invention can be further improved as follows:
further, the data processing control unit is further configured to:
and screening key frames of the point cloud data, taking the key frames as nodes for graph optimization, adding pre-integration information of an inertial navigation unit as constraint information between the nodes, updating the pose after graph optimization constraint and screening constraint are established, and updating the octree map according to the updated pose information.
Further, the screening constraints include:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
tracking less than 50 map point clouds by the current frame, wherein a certain difference is required between the key frame and the map;
the key frames themselves have sufficient and uniformly distributed feature points.
Further, the graph optimization constraints include:
pre-integration information of the inertial navigation unit;
reducing the cumulative error.
A construction method of an octree map comprises the following steps:
s101, point cloud data of the surrounding environment are obtained through a laser radar sensor;
s102, acquiring acceleration information and angular velocity information of equipment through an inertial navigation unit;
s103, performing pre-integration on the acceleration information and the angular velocity information through a data processing control unit to obtain current pose information of the laser radar sensor;
s104, distortion correction is carried out on the point cloud data through the current pose information;
s105, detecting characteristic points of data in the point cloud by a curvature method, and calculating pose transformation information of the laser radar sensor and position information of the current point cloud in space by using the characteristic points for scene correlation;
and S106, creating and updating an octree map according to the pose transformation information and the position information.
Further, the updating the octree map in S106 specifically includes:
s1061, screening key frames of the point cloud data, taking the key frames as nodes of graph optimization, adding pre-integration information of an inertial navigation unit as constraint information between the nodes, updating the pose after graph optimization constraint and screening constraint are established, and updating the octree map according to the updated pose information.
Further, the screening constraint in S1061 specifically includes:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
the current frame tracks less than 50 map point clouds, and a certain difference is needed between the key frame and the map;
the key frame characteristic points are sufficient and uniformly distributed.
Further, the graph optimization constraint in S1061 specifically includes:
obtaining pose transformation information of the laser radar sensor in a certain time period through pre-integration information of the inertial navigation unit;
the accumulated error is reduced by calculating the pose of the current frame and the frame with less accumulated error, and the accumulated error is reduced by reducing the constraint number.
An apparatus for construction of an octree map, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, implementing the steps of the method of constructing an octree map.
An electronic device having stored thereon an implementation program for information transfer, which when executed by a processor implements the steps of a method of constructing an octree map.
The invention has the following advantages:
the construction system of the octree map performs scene association according to the characteristic points in the point cloud, calculates the position of the current point cloud in space, and constructs and updates the octree map. And updating the map and the pose through a loop detection method of the octree map and pre-integral information of the inertial navigation unit. The method combining the inertial navigation unit and the laser ranging sensor can remove the distortion problem of point cloud data obtained by the laser radar under the condition of high-speed movement or large rotation amplitude, can eliminate accumulated errors of pose estimation in the image optimization process, extracts characteristic points in the point cloud for scene association, and uses the obtained sensor pose for updating the octree map, thereby ensuring the accuracy and efficiency of scene association, and creating and updating the octree map to eliminate the influence of dynamic objects. In addition, the created octree map can be used for subsequent robot navigation; the problem of among the prior art utilize the octree map to estimate the relatively poor precision of the motion of point cloud is solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic control diagram of a system for constructing an octree map according to an embodiment of the present invention;
FIG. 2 is a detailed flowchart of a method for constructing an octree map according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an apparatus for constructing an octree map according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a method for constructing an octree map according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating a storage format of an octree map according to an embodiment of the present invention.
The system comprises a laser radar sensor 10, an inertial navigation unit 20 and a data processing control unit 30.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, an octree map building system includes:
a laser radar sensor 10 for acquiring point cloud data of a surrounding environment;
an inertial navigation unit 20 for acquiring acceleration information and angular velocity information of the apparatus;
the data processing control unit 30 is connected with the laser radar sensor 10 and the inertial navigation unit 20, performs pre-integration on acceleration information and angular velocity information to obtain current pose information of the laser radar sensor 10, and performs distortion correction on point cloud data through the current pose information;
detecting the feature points of the data in the point cloud by a curvature method, calculating the pose transformation information of the laser radar sensor 10 and the position information of the current point cloud in the space by using the feature points for scene correlation, and creating and updating an octree map according to the pose transformation information and the position information.
The created map is different from the traditional characteristic map, an octree map of a three-dimensional space is constructed and updated in real time by using information associated with an inertial navigation unit and characteristics, moving objects can be eliminated in the process of map construction, and in addition, the obtained map can be effectively used for navigation of a robot.
As shown in FIG. 3, the construction system of the octree map uses a handheld device as a carrying platform, and comprises a power supply module, a router (used for sensor data transmission) and the like, the laser radar sensor 10 is installed on the top end of the device, a better visual angle is provided for conveniently detecting the surrounding environment, the sensing range of the laser radar sensor 10 is 360 degrees in the horizontal direction, and the sensing range of the laser radar sensor is +/-15 degrees in the vertical direction. The octree map construction system uses a notebook computer as a data processing unit of a sensor, wherein the inertial navigation unit 20 provides information such as attitude, distance and the like for autonomous navigation and positioning of the robot and map construction.
As shown in fig. 4, the data is preprocessed and a local feature point map is constructed:
(a) the handheld mapping device transmits the ambient point cloud frame acquired by the laser radar sensor 10 and the device pose transformation information acquired by the inertial navigation unit through the router. The notebook computer performs pre-integration on inertial navigation unit data through an improved SLAM algorithm, and performs distortion correction on point cloud data by using an obtained result; and calculating curvature information of the points, screening geometric feature points with more information in sequence, estimating pose transformation of the sensor through the feature points and a least square algorithm, and constructing a local feature point map.
(b) Building and updating octree map
An octree map models a three-dimensional space into a number of voxel nodes. In an octree, information whether it is occupied or not is stored in a node. When all children nodes of a block are occupied or not, it is not necessary to expand the node. On the basis, the storage efficiency of the octree map is greatly improved. When an octree map is created, scene association is carried out by adopting the characteristic points, pose transformation of a robot is solved, the point cloud is converted into a coordinate system of a real space by utilizing the result, and the occupation probability of the octree map is updated by a grid map updating method based on a two-dimensional space. Furthermore, during the update, due to the sequence of frames in which the dynamic object exists, the dynamic object will be filtered out in the octree map over time.
The map optimization and map update process:
the pose estimation is usually a recursive process, that is, the pose of the current frame is solved by the pose of the previous frame, and the process of accumulating the error of the pose estimation of the previous frame to the next frame exists, so that the accumulated error is continuously increased along with the time and the enlargement of the mapping scene. To reduce this error, we add back-end graph optimization to the front-end inter-frame correlation to eliminate the accumulated error. The pose is constrained and optimized mainly through data of an inertial navigation unit, a loop frame in the process of establishing the image and the like.
Point cloud distortion correction:
because the data acquired by the laser radar are not at the same moment, the acquired point cloud data have distortion under the condition of high-speed movement or large angle transformation. Compared with a laser radar, the output frequency of the inertial navigation unit is very high, usually at 100-1kHz, the high-frequency acceleration and the angular velocity obtained by the inertial navigation unit are converted into pose transformation, and distorted point cloud can be corrected.
Local feature point map construction:
the characteristic points are often point cloud parts with more information content in the point cloud frame, and the efficiency and the precision of scene association can be improved by detecting the point cloud parts and using the point cloud parts for scene association. The method comprises the steps of extracting features of point clouds by using a curvature algorithm used by a radar odometer, firstly calculating the curvature of each point in the point clouds, and sequencing the curvature, wherein the larger curvature means that the change amplitude of the point cloud is larger, and the point cloud is used as an angle special diagnosis point, and the smaller curvature means that the change amplitude of an area adjacent to the point is smaller, and the point cloud is used as a surface feature point.
After the points are classified, the pose transformation of the sensor is estimated using least squares optimization. Least squares optimization estimates pose motion by minimizing the distance between previous and subsequent frames. For the corner point characteristics, the Euclidean distance between points is used as the least square optimization distance; for line feature points, using the point-to-line distance as the least squares optimization distance; for the face feature points, the point-face distance is used as the least squares optimization distance.
As shown in fig. 5, construction of an octree map:
octrees are hierarchical data structures subdivided in three dimensions. Octree maps, which are octree-based, estimate whether the current space is occupied using probabilistic occupancy, by compressing the data using a structure of trees in a data structure, each node in the octree representing a space contained in a cubic volume, commonly referred to as a voxel. This volume is recursively subdivided into eight sub-volumes until a given minimum voxel size is reached. This compression method ensures compactness of the final model and the octree enables fast retrieval of the state of the space. Because the octree is a hierarchical data structure, the tree can be cut at any level to obtain coarser map nodes.
Before creating the octree map, firstly, the pose of the sensor is transformed through the feature points of the point cloud, and the current frame point cloud is mapped to the actual space by utilizing the transformed pose. The space is divided into data blocks by setting the size of the voxel, and the occupation probability of the current voxel is updated according to the current frame point cloud position and the existing octree map, so that the octree map is created.
Unlike the superposition of points on the map used by the point cloud map, the octree map continuously updates the occupancy probability of the current voxel during the creation process, so for moving objects, the number of times they appear in the data frame is often small, and for moving objects, the occupancy probability is continuously reduced over time.
Updating the octree map:
because the pose is predicted by adopting a scene correlation method between the previous frame and the next frame, the problem exists in that the predicted error of the previous frame can be accumulated in the next frame, and in order to eliminate the influence, a graph optimization method is introduced to optimize the pose.
The pose of the robot is used as a node for graph optimization, the pre-integral information of an inertial navigation unit is added as constraint information between the nodes, in addition, due to the complexity of the environment, the robot tends to return to a place passing by before in the graph building process, the accumulation of errors can be reduced by calculating the pose of a current frame and a frame with less accumulated errors, and the pose constraint relation is called a loop. By reducing the number of constraints, the cumulative error is reduced. The method comprises the steps of utilizing the advantages of an octree map to detect a loop, converting a current frame into an octree map form, utilizing pose transformation to find a local map with close distance, carrying out probability matching on octree voxels of the current frame and the local map, finally obtaining the pose of the loop frame, and adding the constraint into map optimization to optimize the pose.
On the basis of the technical scheme, the invention can be further improved as follows:
further, the data processing control unit 30 is further configured to:
and screening key frames of the point cloud data, taking the key frames as nodes for graph optimization, adding pre-integration information of an inertial navigation unit as constraint information between the nodes, updating the pose after graph optimization constraint and screening constraint are established, and updating the octree map according to the updated pose information.
Further, the screening constraints include:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
tracking less than 50 map point clouds by the current frame, wherein a certain difference is required between the key frame and the map;
the key frames themselves have sufficient and uniformly distributed feature points.
Further, the graph optimization constraints include: obtaining pose transformation information of the laser radar sensor 10 in a certain time period through pre-integration information of the inertial navigation unit;
the accumulated error is reduced by calculating the pose of the current frame and the frame with less accumulated error, and the accumulated error is reduced by reducing the constraint number.
As shown in fig. 2, a method for constructing an octree map specifically includes:
s101, point cloud data of the surrounding environment are obtained;
in the step, point cloud data of the surrounding environment is obtained through a laser radar sensor 10;
s102, acquiring acceleration information and angular velocity information of equipment;
in this step, the inertial navigation unit 20 obtains acceleration information and angular velocity information of the device;
s103, obtaining current pose information of the laser radar sensor 10;
in this step, the data processing control unit 30 performs pre-integration on the acceleration information and the angular velocity information to obtain the current pose information of the laser radar sensor 10;
s104, carrying out distortion correction on the point cloud data;
in the step, distortion correction is carried out on point cloud data through current pose information;
s105, calculating pose transformation information and position information of a current point;
in the step, feature points of data in the point cloud are detected through a curvature method, and scene correlation is carried out by using the feature points to calculate the position and orientation transformation information of the laser radar sensor 10 and the position information of the current point cloud in the space;
and S106, creating and updating the octree map.
In this step, an octree map is created and updated according to the pose transformation information and the position information.
Further, S1061, updating the octree map according to the updated pose information.
In the step, a key frame of the point cloud data is screened and used as a node of the graph optimization, pre-integral information of an inertial navigation unit is added as constraint information between the nodes, after the graph optimization constraint and the screening constraint are established, the pose is updated, and the octree map is updated according to the updated pose information.
Further, the screening constraint in S1061 specifically includes:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
the current frame tracks less than 50 map point clouds, and a certain difference is needed between the key frame and the map;
the key frame characteristic points are sufficient and uniformly distributed.
Further, the graph optimization constraint in S1061 specifically includes:
obtaining pose transformation information of the laser radar sensor 10 in a certain time period through pre-integration information of the inertial navigation unit;
the accumulated error is reduced by calculating the pose of the current frame and the frame with less accumulated error, and the accumulated error is reduced by reducing the constraint number.
An apparatus for construction of an octree map, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, implementing the steps of the method of constructing an octree map.
An electronic device having stored thereon an implementation program for information transfer, which when executed by a processor implements the steps of a method of constructing an octree map.
The construction system for the octree map comprises the following steps:
when the device is used, point cloud data of the surrounding environment are obtained through the laser radar sensor 10; acquiring acceleration information and angular velocity information of the device through the inertial navigation unit 20; pre-integrating the acceleration information and the angular velocity information through a data processing control unit 30 to obtain the current pose information of the laser radar sensor 10; carrying out distortion correction on point cloud data through current pose information; detecting characteristic points of data in the point cloud by a curvature method, and calculating pose transformation information of the laser radar sensor 10 and position information of the current point cloud in space by using the characteristic points to perform scene correlation; and creating and updating an octree map according to the pose transformation information and the position information.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
One or more embodiments of the present description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more embodiments of the specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only an example of this document and is not intended to limit this document. Various modifications and changes may occur to those skilled in the art from this document. Any modifications, equivalents, improvements, etc. which come within the spirit and principle of the disclosure are intended to be included within the scope of the claims of this document.
Claims (10)
1. An octree map construction system, comprising:
the system comprises a laser radar sensor, a data processing unit and a data processing unit, wherein the laser radar sensor is used for acquiring point cloud data of the surrounding environment;
the inertial navigation unit is used for acquiring acceleration information and angular velocity information of the equipment;
the data processing control unit is connected with the laser radar sensor and the inertial navigation unit, performs pre-integration on acceleration information and angular velocity information to obtain current pose information of the laser radar sensor, and performs distortion correction on point cloud data through the current pose information; detecting feature points of data in the point cloud by a curvature method, utilizing the feature points to perform scene correlation calculation on pose transformation information of the laser radar sensor and position information of the current point cloud in space, and creating and updating an octree map according to the pose transformation information and the position information.
2. The octree map building system according to claim 1, wherein the data processing control unit is further configured to:
and screening key frames of the point cloud data, taking the key frames as nodes for graph optimization, adding pre-integration information of an inertial navigation unit as constraint information between the nodes, updating the pose after graph optimization constraint and screening constraint are established, and updating the octree map according to the updated pose information.
3. The octree map building system according to claim 2, wherein the filtering constraints comprise:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
the current frame tracks less than 50 map point clouds, and a certain difference is needed between the key frame and the map;
the key frame characteristic points are sufficient and uniformly distributed.
4. The octree map building system according to claim 2, wherein the graph optimization constraints comprise:
pre-integration information of the inertial navigation unit;
reducing the cumulative error.
5. A construction method of an octree map is characterized by specifically comprising the following steps:
s101, point cloud data of the surrounding environment are obtained through a laser radar sensor;
s102, acquiring acceleration information and angular velocity information of equipment through an inertial navigation unit;
s103, performing pre-integration on the acceleration information and the angular velocity information through a data processing control unit to obtain current pose information of the laser radar sensor;
s104, distortion correction is carried out on the point cloud data through the current pose information;
s105, detecting characteristic points of data in the point cloud by a curvature method, and calculating pose transformation information of the laser radar sensor and position information of the current point cloud in space by using the characteristic points for scene correlation;
and S106, creating and updating an octree map according to the pose transformation information and the position information.
6. The method for constructing an octree map according to claim 5, wherein the octree map in S106 specifically comprises:
s1061, screening key frames of the point cloud data, taking the key frames as nodes of graph optimization, adding pre-integration information of an inertial navigation unit as constraint information between the nodes, updating the pose after graph optimization constraint and screening constraint are established, and updating the octree map according to the updated pose information.
7. The method for constructing an octree map according to claim 6, wherein the filtering constraints in S1061 specifically include:
a certain distance is needed between two key frames;
the current frame tracks less than 75% of the reference key frame;
the current frame tracks less than 50 map point clouds, and a certain difference is needed between the key frame and the map;
the key frame characteristic points are sufficient and uniformly distributed.
8. The method for constructing an octree map according to claim 7, wherein the graph optimization constraints in S1061 specifically include:
obtaining pose transformation information of the laser radar sensor in a certain time period through pre-integration information of the inertial navigation unit;
the accumulated error is reduced by calculating the pose of the current frame and the frame with less accumulated error, and the accumulated error is reduced by reducing the constraint number.
9. An apparatus for construction of an octree map, comprising: memory, processor and computer program stored on the memory and executable on the processor, which computer program, when executed by the processor, carries out the steps of the method of constructing an octree map according to any one of claims 5 to 8.
10. An electronic device, characterized in that the electronic device stores an implementation program for information transfer, and the program is executed by a processor to implement the steps of the octree map construction method according to any one of claims 5 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111354683.7A CN114037800A (en) | 2021-11-16 | 2021-11-16 | Construction system, method and device of octree map and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111354683.7A CN114037800A (en) | 2021-11-16 | 2021-11-16 | Construction system, method and device of octree map and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114037800A true CN114037800A (en) | 2022-02-11 |
Family
ID=80144493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111354683.7A Pending CN114037800A (en) | 2021-11-16 | 2021-11-16 | Construction system, method and device of octree map and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114037800A (en) |
-
2021
- 2021-11-16 CN CN202111354683.7A patent/CN114037800A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112014857B (en) | Three-dimensional laser radar positioning and navigation method for intelligent inspection and inspection robot | |
| CN112347840B (en) | Vision sensor laser radar integrated unmanned aerial vehicle positioning and image building device and method | |
| CN111220993B (en) | Target scene positioning method and device, computer equipment and storage medium | |
| EP2849117B1 (en) | Methods, apparatuses and computer program products for automatic, non-parametric, non-iterative three dimensional geographic modeling | |
| CN113781582A (en) | Synchronous positioning and map creating method based on laser radar and inertial navigation combined calibration | |
| CN108267141B (en) | Road point cloud data processing system | |
| CN102831646A (en) | Scanning laser based large-scale three-dimensional terrain modeling method | |
| CN110298914B (en) | Method for establishing fruit tree canopy feature map in orchard | |
| CN113593017A (en) | Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium | |
| CN113269837A (en) | Positioning navigation method suitable for complex three-dimensional environment | |
| CN111862214B (en) | Computer equipment positioning method, device, computer equipment and storage medium | |
| CN115388902B (en) | Indoor positioning method and system, AR indoor positioning navigation method and system | |
| CN115479598A (en) | Positioning and mapping method based on multi-sensor fusion and tight coupling system | |
| CN111768489B (en) | Indoor navigation map construction method and system | |
| CN106997614A (en) | A kind of large scale scene 3D modeling method and its device based on depth camera | |
| CN113218408B (en) | 2Dslam method and system suitable for multi-sensor fusion of multiple terrains | |
| JP4568845B2 (en) | Change area recognition device | |
| CN115900710A (en) | Dynamic environment navigation method based on visual information | |
| CN111207753A (en) | Method for simultaneously positioning and establishing picture under multi-glass partition environment | |
| CN114677435A (en) | Point cloud panoramic fusion element extraction method and system | |
| CN114429432B (en) | Multi-source information layered fusion method and device and storage medium | |
| CN115183762A (en) | Airport warehouse inside and outside mapping method, system, electronic equipment and medium | |
| CN115355904A (en) | Slam method for Lidar-IMU fusion of ground mobile robot | |
| CN116429116A (en) | Robot positioning method and equipment | |
| CN113838129B (en) | Method, device and system for obtaining pose information |
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 |