CN113961013A - Unmanned aerial vehicle path planning method based on RGB-D SLAM - Google Patents
Unmanned aerial vehicle path planning method based on RGB-D SLAM Download PDFInfo
- Publication number
- CN113961013A CN113961013A CN202111337558.5A CN202111337558A CN113961013A CN 113961013 A CN113961013 A CN 113961013A CN 202111337558 A CN202111337558 A CN 202111337558A CN 113961013 A CN113961013 A CN 113961013A
- Authority
- CN
- China
- Prior art keywords
- aerial vehicle
- unmanned aerial
- point
- path planning
- rgb
- 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
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000010276 construction Methods 0.000 claims abstract description 6
- 239000013598 vector Substances 0.000 claims description 12
- 108010046685 Rho Factor Proteins 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000011161 development Methods 0.000 abstract description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
- G05D1/0808—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
- G05D1/106—Change initiated in response to external conditions, e.g. avoidance of elevated terrain or of no-fly zones
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides an unmanned aerial vehicle path planning method based on RGB-D SLAM, and particularly relates to the field of multi-rotor unmanned aerial vehicle path planning. The specific method comprises the following steps: the construction of a field environment map is completed by the front end by utilizing an RGB-D SLAM technology, an artificial potential field path planning algorithm is operated after an initial coordinate point and a terminal coordinate point are given until a path planning task is completed, and if the problem of falling into a local minimum value and the like occurs, a fault processing program is entered, so that the working reliability of the unmanned aerial vehicle is improved. The path planning method provided by the invention combines the RGB-D SLAM technology and the path planning algorithm, has a simple and reliable structure, can construct an environment map by itself, greatly improves the flying working efficiency of the unmanned aerial vehicle, and has important significance for the development of the autonomous navigation technology of the unmanned aerial vehicle.
Description
Technical Field
The invention relates to the field of multi-rotor unmanned aerial vehicle path planning, in particular to an unmanned aerial vehicle path planning method based on RGB-DSLAM.
Background
With the heat of artificial intelligence, the intelligent mobile robot technology has been rapidly developed and widely applied to the fields of military, traffic, home service and the like. In China, Baidu Apollo unmanned vehicles, American groups automatic delivery vehicles and the like have been taken first for trial operation, and various automatic mobile service robots are gradually civilized. The most key technology for realizing the functions is to solve the technical problem of path planning, so that the automatic cruise driving function is realized.
And in the unmanned aerial vehicle automatic cruise field, owing to be subject to unmanned aerial vehicle's load capacity, battery continuation of the journey and the frequency of processing of carrying on the treater, unmanned aerial vehicle's the sensor that carries on needs miniaturization, lightweight as far as, otherwise can influence unmanned aerial vehicle self flight hang time.
At present, the path planning task of the civil unmanned aerial vehicle is mainly realized on the basis of GPS + IMU, although the technical scheme is mature and widely applied, under the conditions of GPS signal loss and complex field environment, the unmanned aerial vehicle cannot complete path planning calculation of a target point and cannot autonomously reach a specified place.
Disclosure of Invention
In order to overcome the defects, the invention provides an unmanned aerial vehicle path planning method based on RGB-D SLAM, and simultaneously completes the real-time construction of an unmanned aerial vehicle front-end map and the fusion design of a rear-end path planning algorithm. The real-time autonomous navigation function of the unmanned aerial vehicle can be realized in a complex environment.
The technical scheme adopted by the invention is as follows:
an unmanned aerial vehicle path planning method based on RGB-D SLAM is characterized in that: the method comprises the following steps:
s1, constructing an unmanned aerial vehicle three-dimensional point cloud map by the RGB-D SLAM system, and realizing accurate positioning of the unmanned aerial vehicle.
And S2, realizing the real-time path planning capability of the unmanned aerial vehicle by adopting an artificial potential field method.
S3, designing an exception handling program of a path planning algorithm, and greatly improving the response capability of the unmanned aerial vehicle in actually handling the fault problem.
Further, the RGB-D SLAM system of step S1 is composed of an RGB-D sensor and a mobile PC, and the RGB-D visual sensor with lower cost is used to acquire a depth color image, and the mobile PC is used to process a large amount of image data streams, and the system is a map front-end system of the unmanned aerial vehicle. The kernel algorithm adopted by the RGB-D SLAM is an ORB-SLAM scheme, and the specific flow is as follows:
s101, ORB feature extraction is carried out on the color data image obtained by RGB-D, and initial matching is carried out on the color data image and the image feature point of the previous frame in order to obtain the initial point of the feature. And meanwhile, depth calculation is carried out on the image, and a depth image is further obtained.
And S102, judging the number of the feature points of the data stream obtained by the last number, and if the feature matching points of the depth images between the front frame and the rear frame are larger than a set value, estimating the poses of the front frame and the rear frame by using an RANSAC algorithm.
And S103, if the matching number of the feature points is smaller than a set value in the step S102, turning to calculating bag-of-word vectors of the current frame, wherein the bag-of-word vectors of the system are obtained by establishing ORB features of the key frame by adopting a DBoW2 library. And searching and matching the bag-of-words vector of the current frame with the key frame data, and then performing optimal pose estimation between the current frame and each key frame by adopting an RANSAC algorithm.
And S104, performing current frame attitude optimization on the data streams obtained in S102 and S103 by using the local map obtained by the collaborative map, further obtaining information such as the pose, the feature point and the camera parameter of the current frame, and then judging the key frame.
And S105, sending a part of the data processed in the step S104 into a bag-of-words model for searching and matching the bag-of-words vector in the step S103 and the key frame data, adopting g2o to perform BA adjustment on the processed key frame and the related key frame at different internal moments, and finally performing closed-loop detection.
And S106, processing the results obtained in the steps through an octree data structure, and integrating the unordered point cloud data stream into indexable ordered 3D point cloud data, so that the construction of the front-end map system of the unmanned aerial vehicle autonomous navigation is completed.
Further, the path planning method in step S2 adopts an artificial potential field method, and the algorithm flow specifically includes the following steps:
s201, determining a position X ═ X, y, z in the current state space of operation of the drone]TAnd x, y and z are space three-dimensional coordinates, and a gravitational field function is introduced into the space three-dimensional coordinates:
where ε is the gravitational scale factor, ρ (q, q)goal) For unmanned aerial vehicle q and target point qgoalIs a Euclidean distance, Uatt(q) is the function value of the gravitational field.
S202, introducing a potential force function to the potential field:
wherein FattAnd (q) is a gravitational field gravitational value.
Potential field function U corresponding to repulsive forcereq(q) is:
where eta is a repulsive force scale factor, ρ (q, q)obs) For robot q and obstacle qobsEuropean distance between them, p0Is the range of influence of the obstacle. The repulsion function can then be expressed as:
wherein:
in the above formula Frep1(q) is the component of motion in the potential force field that repels the robot away from the obstacle, Frep2(q) is the motion component of the target point attracting the robot in the potential force field.
S203, further obtaining a force function F in the three-dimensional state point cloud of the unmanned aerial vehicle from S201 to S202sum(q) is:
further, the exception handling procedure in step S3 is to correct various local minimum problems that occur when the unmanned aerial vehicle operates in the artificial potential field method, and in this problem, the unmanned aerial vehicle is likely to get into a certain local minimum point and cannot complete global path planning, and cannot reach a designed target point, and the specific exception handling procedure of the unmanned aerial vehicle is as follows:
s301, setting the radius of a three-dimensional space sphere in a state space as R, the jump point coefficient as alpha, and the hovering threshold time as T, and when the unmanned aerial vehicle reaches a certain point, the resultant external force F of the gravitational field is 0, and the hovering time exceeds the set threshold time T, generating a spherical surface with the radius of alpha R by taking the unmanned aerial vehicle as the center of a circle.
S302, randomly selecting a point on the spherical surface, judging whether the point is positioned on the obstacle, if so, deleting the point, then randomly selecting a point, continuously and repeatedly judging whether the point is positioned on the obstacle, and limiting the traversal time TsearchA position point is found internally and set as a local target point D of the current transfer, and the unmanned aerial vehicle operates the artificial potential field method to fly toAnd after the point, continuing to operate the artificial potential field method to the target point until the final path planning task is completed.
S303, if T is in step S302searchIf no local target point is found, generating a fault report error, and selecting two working modes: the first mode is as follows: inputting a new return flight terminal coordinate value and executing an artificial potential field algorithm to carry out return flight. And a second mode: the current position point is dropped. Wherein T issearchThe set value of (a) is adjusted according to the computing power of the PC, and the jump point coefficient α is set according to the accuracy of the flight map.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
1. the invention provides an unmanned aerial vehicle path planning method based on RGB-D SLAM, provides a new path planning algorithm, and adopts a low-cost technical scheme to realize a relatively reliable three-dimensional field environment map
2. The unmanned aerial vehicle path planning method based on the RGB-D SLAM can realize autonomous scene map construction and autonomous path planning in a complex field environment, effectively broadens the application occasions of autonomous operation of the unmanned aerial vehicle, and has important significance for the development of the unmanned aerial vehicle autonomous navigation technology.
Drawings
FIG. 1: is a work flow diagram of the present invention;
FIG. 2: is a flow chart of the fault procedure of the present invention.
Detailed Description
The following further describes embodiments of the present invention with reference to the accompanying drawings:
the invention relates to an unmanned aerial vehicle path planning method based on RGB-D SLAM, and a work flow chart is shown in figure 1.
The specific embodiment of the invention is as follows:
s1, constructing an unmanned aerial vehicle three-dimensional point cloud map by the RGB-D SLAM system, and realizing accurate positioning of the unmanned aerial vehicle.
And S2, realizing the real-time path planning capability of the unmanned aerial vehicle by adopting an artificial potential field method.
S3, designing an exception handling program of a path planning algorithm, and greatly improving the response capability of the unmanned aerial vehicle in actually handling the fault problem.
Further, the RGB-D SLAM system of step S1 is composed of an RGB-D sensor and a mobile PC, and the RGB-D visual sensor with lower cost is used to acquire a depth color image, and the mobile PC is used to process a large amount of image data streams, and the system is a map front-end system of the unmanned aerial vehicle. The kernel algorithm adopted by the RGB-D SLAM is an ORB-SLAM scheme, and the specific flow is as follows:
s101, ORB feature extraction is carried out on the color data image obtained by RGB-D, and initial matching is carried out on the color data image and the image feature point of the previous frame in order to obtain the initial point of the feature. And meanwhile, depth calculation is carried out on the image, and a depth image is further obtained.
And S102, judging the number of the feature points of the data stream obtained by the last number, and if the feature matching points of the depth images between the front frame and the rear frame are larger than a set value, estimating the poses of the front frame and the rear frame by using an RANSAC algorithm.
And S103, if the matching number of the feature points is smaller than a set value in the step S102, turning to calculating bag-of-word vectors of the current frame, wherein the bag-of-word vectors of the system are obtained by establishing ORB features of the key frame by adopting a DBoW2 library. And searching and matching the bag-of-words vector of the current frame with the key frame data, and then performing optimal pose estimation between the current frame and each key frame by adopting an RANSAC algorithm.
And S104, performing current frame attitude optimization on the data streams obtained in S102 and S103 by using the local map obtained by the collaborative map, further obtaining information such as the pose, the feature point and the camera parameter of the current frame, and then judging the key frame.
And S105, sending a part of the data processed in the step S104 into a bag-of-words model for searching and matching the bag-of-words vector in the step S103 and the key frame data, adopting g2o to perform BA adjustment on the key frame currently processed and the related key frame at different internal moments, and finally performing closed-loop detection.
And S106, processing the results obtained in the steps through an octree data structure, and integrating the unordered point cloud data stream into indexable ordered 3D point cloud data, so that the construction of the front-end map system of the unmanned aerial vehicle autonomous navigation is completed.
Further, the path planning method in step S2 adopts an artificial potential field method, and the algorithm flow specifically includes the following steps:
s201, determining a position X ═ X, y, z in the current state space of operation of the drone]TAnd x, y and z are the left side of the three-dimensional space, and a gravitational field function is introduced:
where ε is the gravitational scale factor, ρ (q, q)goal) For unmanned aerial vehicle q and target point qgoalThe Euclidean distance of (a) is,
s202, introducing a potential force function to the potential field:
wherein FattAnd (q) is a gravitational field gravitational value.
Potential field function U corresponding to repulsive forcereq(q) is:
where eta is a repulsive force scale factor, ρ (q, q)obs) For robot q and obstacle qobsEuropean distance between them, p0Is the range of influence of the obstacle. The repulsion function can then be expressed as:
wherein:
in the above formula Frep1(q) is the component of motion in the potential force field that repels the robot away from the obstacle, Frep2(q) is the motion component of the target point attracting the robot in the potential force field.
S203, further obtaining a force function F in the three-dimensional state point cloud of the unmanned aerial vehicle from S201 to S202sum(q) is:
further, as shown in fig. 2, the exception handling procedure in step S3 is to correct various local minimum problems occurring when the unmanned aerial vehicle operates the artificial potential field method, and in this problem, the unmanned aerial vehicle is likely to fall into a certain local minimum point and cannot complete global path planning, and cannot reach a designed target point, and the specific exception handling procedure of the unmanned aerial vehicle is as follows:
s301, setting the radius of a three-dimensional space sphere in a state space as R, the jump point coefficient as alpha, and the hovering threshold time as T, and when the unmanned aerial vehicle reaches a certain point, the resultant external force F of the gravitational field is 0, and the hovering time exceeds the set threshold time T, generating a spherical surface with the radius of alpha R by taking the unmanned aerial vehicle as the center of a circle.
S302, randomly selecting a point on the spherical surface, judging whether the point is positioned on the obstacle, if so, deleting the point, then randomly selecting a point, continuously and repeatedly judging whether the point is positioned on the obstacle, and limiting the traversal time TsearchAnd searching a position point, setting the position point as a current transfer local target point D, and continuing to operate the artificial potential field method to the target point after the unmanned aerial vehicle operates the artificial potential field method to fly to the position point until a final path planning task is completed.
S303, if T is in step S302searchIf no local target point is found, generating fault report and selecting two working modes: the first mode is as follows: inputting a new return flight terminal coordinate value and executing an artificial potential field algorithm to carry out return flight. And a second mode: the current position point is dropped. Wherein T issearchThe set value of (a) is adjusted according to the computing power of the PC, and the jump point coefficient α is set according to the accuracy of the flight map.
Claims (4)
1. An unmanned aerial vehicle path planning method based on RGB-D SLAM is characterized in that:
s1, constructing an unmanned aerial vehicle three-dimensional point cloud map by the RGB-D SLAM system to realize accurate positioning of the unmanned aerial vehicle;
s2, realizing the real-time path planning capability of the unmanned aerial vehicle by adopting an artificial potential field method;
s3, designing an exception handling program of a path planning algorithm, and greatly improving the response capability of the unmanned aerial vehicle in actually handling the fault problem.
2. The method of claim 1, wherein the method comprises: the RGB-D SLAM system of step S1 is composed of an RGB-D sensor and a mobile PC, and uses the RGB-D vision sensor with lower cost to obtain a depth color image, and uses the mobile PC to process a large amount of image data streams, and the system is a map front-end system of the unmanned aerial vehicle; the kernel algorithm adopted by the RGB-D SLAM is an ORB-SLAM scheme, and the specific flow is as follows:
s101, performing ORB feature extraction on a color data image obtained by RGB-D, and performing initial matching on an initial point of a feature and an image feature point of a previous frame; meanwhile, depth calculation is carried out on the image, and a depth image is obtained;
s102, judging the number of feature points of the data stream obtained by the last number, and if the feature matching points of the depth image between the front frame and the rear frame are larger than a set value, estimating the poses of the front frame and the rear frame by using an RANSAC algorithm;
s103, if the matching number of the feature points is smaller than a set value in the step S102, turning to calculation of bag-of-word vectors of the current frame, wherein the bag-of-word vectors of the system are obtained by establishing ORB features of the key frame by adopting a DBoW2 library; searching and matching the bag-of-word vector of the current frame with the key frame data, and then performing optimal pose estimation between the current frame and each key frame by adopting an RANSAC algorithm;
s104, performing current frame attitude optimization on the data streams obtained in S102 and S103 by using a local map obtained by the collaborative map, further obtaining information such as the pose, the feature points and the camera parameters of the current frame, and then judging the key frame;
s105, sending a part of the data processed in the step S104 into a bag-of-words model for searching and matching the bag-of-words vector in the step S103 and the key frame data, adopting g2o to perform BA adjustment on the key frame currently processed and the related key frames at different internal moments, and finally performing closed-loop detection;
and S106, processing the results obtained in the steps through an octree data structure, and integrating the unordered point cloud data stream into indexable ordered 3D point cloud data, so that the construction of the front-end map system of the unmanned aerial vehicle autonomous navigation is completed.
3. The RGB-D SLAM-based unmanned aerial vehicle path planning method of claim 2, wherein: the path planning method in the step S2 adopts an artificial potential field method, and the algorithm flow is specifically as follows:
s201, determining a position X ═ X, y, z in the current state space of operation of the drone]TAnd x, y and z are space three-dimensional coordinates, and a gravitational field function is introduced into the space three-dimensional coordinates:
where ε is the gravitational scale factor, ρ (q, q)goal) For unmanned aerial vehicle q and target point qgoalIs a Euclidean distance, Uatt(q) is a function value of the gravitational field;
s202, introducing a potential force function to the potential field:
wherein Fatt(q) is a gravitational field gravitational value;
potential field function U corresponding to repulsive forcereq(q) is:
where eta is a repulsive force scale factor, ρ (q, q)obs) For unmanned plane q and obstacle qobsEuropean distance between them, p0Is the influence range of the obstacle; the repulsion function can then be expressed as:
wherein:
in the above formula Frep1(q) is the component of motion in the potential force field that repels the drone away from the obstacle, Frep2(q) is a motion component of a target point in the potential force field attracting the unmanned aerial vehicle;
s203, further obtaining a force function in the unmanned aerial vehicle three-dimensional state point cloud from S201 to S202 as follows:
4. the RGB-D SLAM-based unmanned aerial vehicle path planning method of claim 3, wherein: the exception handling procedure in step S3 is to correct various local minimum problems that occur when the unmanned aerial vehicle operates the artificial potential field method, and in this case, the unmanned aerial vehicle is likely to get into a certain local minimum point and cannot complete global path planning, and cannot reach a designed target point, and the specific exception handling procedure for the unmanned aerial vehicle is as follows:
s301, setting the radius of a three-dimensional space sphere in a state space as R, the jump point coefficient as alpha, and the hovering threshold time as T, and when the unmanned aerial vehicle reaches a certain point, the resultant external force F of the gravitational field is 0 and the hovering time exceeds the set threshold time T, generating a spherical surface with the radius of alpha R by taking the unmanned aerial vehicle as the center of a circle;
s302, randomly selecting a point on the spherical surface, judging whether the point is positioned on the obstacle, if so, deleting the point, then randomly selecting a point, continuously and repeatedly judging whether the point is positioned on the obstacle, and limiting the traversal time TsearchSearching a position point internally, setting the position point as a current transfer local target point D, and continuing to operate the artificial potential field method to the target point after the unmanned aerial vehicle operates the artificial potential field method to fly to the position point until a final path planning task is completed;
s303, if T is in step S302searchIf no local target point is found, generating a fault report error, and selecting two working modes: inputting a new return terminal coordinate value and executing an artificial potential field algorithm to carry out return flight in the first mode; mode two Current position Point Fall, wherein TsearchThe set value of (a) is adjusted according to the computing power of the PC, and the jump point coefficient α is set according to the accuracy of the flight map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111337558.5A CN113961013A (en) | 2021-11-10 | 2021-11-10 | Unmanned aerial vehicle path planning method based on RGB-D SLAM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111337558.5A CN113961013A (en) | 2021-11-10 | 2021-11-10 | Unmanned aerial vehicle path planning method based on RGB-D SLAM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113961013A true CN113961013A (en) | 2022-01-21 |
Family
ID=79470219
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111337558.5A Pending CN113961013A (en) | 2021-11-10 | 2021-11-10 | Unmanned aerial vehicle path planning method based on RGB-D SLAM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113961013A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115061492A (en) * | 2022-06-20 | 2022-09-16 | 华南理工大学 | Campus takeout distribution system and progressive three-dimensional space path planning method |
| CN117346791A (en) * | 2023-12-01 | 2024-01-05 | 北京科技大学 | Intelligent wheelchair path planning method and system based on visual images |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590827A (en) * | 2017-09-15 | 2018-01-16 | 重庆邮电大学 | A kind of indoor mobile robot vision SLAM methods based on Kinect |
| JP2021077353A (en) * | 2019-11-05 | 2021-05-20 | 広東工業大学Ｇｕａｎｇｄｏｎｇ Ｕｎｉｖｅｒｓｉｔｙ Ｏｆ Ｔｅｃｈｎｏｌｏｇｙ | Drone vision slam method based on gpu acceleration |
| CN112904901A (en) * | 2021-01-14 | 2021-06-04 | 吉林大学 | Path planning method based on binocular vision slam and fusion algorithm |
-
2021
- 2021-11-10 CN CN202111337558.5A patent/CN113961013A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590827A (en) * | 2017-09-15 | 2018-01-16 | 重庆邮电大学 | A kind of indoor mobile robot vision SLAM methods based on Kinect |
| JP2021077353A (en) * | 2019-11-05 | 2021-05-20 | 広東工業大学Ｇｕａｎｇｄｏｎｇ Ｕｎｉｖｅｒｓｉｔｙ Ｏｆ Ｔｅｃｈｎｏｌｏｇｙ | Drone vision slam method based on gpu acceleration |
| CN112904901A (en) * | 2021-01-14 | 2021-06-04 | 吉林大学 | Path planning method based on binocular vision slam and fusion algorithm |
Non-Patent Citations (3)
| Title |
|---|
| 李佳俊: "基于RGBD的无人机语义SLAM***原型设计及实现", pages 1 - 80 * |
| 李元 等: "基于 RGB-D 信息的移动机器人 SLAM 和路径规划 方法研究与实现", 智能***学报, vol. 13, no. 3, pages 445 - 451 * |
| 李卫成 等: "基于RGBD的机器人室内SLAM与路径规划***", 工业控制计算机, vol. 29, no. 3, pages 77 - 81 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115061492A (en) * | 2022-06-20 | 2022-09-16 | 华南理工大学 | Campus takeout distribution system and progressive three-dimensional space path planning method |
| CN117346791A (en) * | 2023-12-01 | 2024-01-05 | 北京科技大学 | Intelligent wheelchair path planning method and system based on visual images |
| CN117346791B (en) * | 2023-12-01 | 2024-03-22 | 北京科技大学 | Intelligent wheelchair path planning method and system based on visual images |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zou et al. | A comparative analysis of LiDAR SLAM-based indoor navigation for autonomous vehicles | |
| CN110068335B (en) | Unmanned aerial vehicle cluster real-time positioning method and system under GPS rejection environment | |
| CN109917818B (en) | Collaborative search containment method based on ground robot | |
| CN105759829A (en) | Laser radar-based mini-sized unmanned plane control method and system | |
| CN112734765B (en) | Mobile robot positioning method, system and medium based on fusion of instance segmentation and multiple sensors | |
| CN111982114B (en) | Rescue robot for estimating three-dimensional pose by adopting IMU data fusion | |
| CN113961013A (en) | Unmanned aerial vehicle path planning method based on RGB-D SLAM | |
| WO2023116667A1 (en) | Charging device, and method for controlling mechanical arm to charge | |
| CN110208783B (en) | Intelligent vehicle positioning method based on environment contour | |
| Jia et al. | A Survey of simultaneous localization and mapping for robot | |
| CN111176309A (en) | Multi-unmanned aerial vehicle self-group mutual inductance understanding method based on spherical imaging | |
| CN113591518B (en) | Image processing method, network training method and related equipment | |
| CN114299039B (en) | Robot and collision detection device and method thereof | |
| Zhang et al. | Online ground multitarget geolocation based on 3-D map construction using a UAV platform | |
| CN108646760B (en) | Monocular vision based mobile robot target tracking and platform control system and method | |
| CN116734850A (en) | Unmanned platform reinforcement learning autonomous navigation system and method based on visual input | |
| Chronis et al. | Path planning of autonomous UAVs using reinforcement learning | |
| Cui et al. | Landmark extraction and state estimation for UAV operation in forest | |
| TW202334613A (en) | Method for searching a path by using a three-dimensional reconstructed map | |
| CN113206951A (en) | Real-time electronic image stabilization method based on flapping wing flight system | |
| Szendy et al. | Simultaneous localization and mapping with TurtleBotII | |
| Wu et al. | Integrated navigation algorithm based on vision-inertial extended Kalman filter for low-cost unmanned aerial vehicle | |
| Jhan et al. | Development of low-cost camera-based UAV control strategy | |
| Wang et al. | Research on Key Technologies of Lunar Patrol Perception in Sparse Scenes | |
| Al Baghdadi et al. | Unmanned aerial vehicles and machine learning for detecting objects in real time |
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 |