CN110208819A - A kind of processing method of multiple barrier three-dimensional laser radar data - Google Patents
A kind of processing method of multiple barrier three-dimensional laser radar data Download PDFInfo
- Publication number
- CN110208819A CN110208819A CN201910398906.6A CN201910398906A CN110208819A CN 110208819 A CN110208819 A CN 110208819A CN 201910398906 A CN201910398906 A CN 201910398906A CN 110208819 A CN110208819 A CN 110208819A
- Authority
- CN
- China
- Prior art keywords
- grid
- point
- data
- barrier
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/86—Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
- G01S17/93—Lidar systems specially adapted for specific applications for anti-collision purposes
- G01S17/931—Lidar systems specially adapted for specific applications for anti-collision purposes of land vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Electromagnetism (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of processing methods of multiple barrier three-dimensional laser radar data, including establish grating map, grid filtering, determine that adapting to four steps of threshold value and mesh connectivity labeled clusters carries out data processing.The utility model has the advantages that processing method proposed by the present invention, realizes the denoising to noise spot and scattered point and simplifies processing；Grid map under different distances can be clustered, barrier discrete part in distant place also can preferably cluster together, obtain vehicle front than more complete barrier profile point and relative position information；Data fusion for laser radar and camera is laid the groundwork, so that the automatic collision avoidance for Active collision avoidance system provides necessary foundation.
Description
Technical field
The present invention relates to a kind of processing method of laser radar data, in particular to a kind of multiple barrier three-dimensional laser thunders
Up to the processing method of data, belong to spatial information data process field.
Background technique
With artificial intelligence, the continuous development of sensor technology, intelligent driving automobile has been made significant headway.Environment sense
Know the basis as intelligent driving, can the superiority and inferiority of performance be directly related to the demand that reach safe driving.Environment sensing is normal
Sensor has visual sensor, laser sensor, inertial navigation, GPS positioning system etc..Laser radar is due to its precision
Height, it is affected by environment smaller, it is widely applied in pilotless automobile.
The three dimensional point cloud that laser radar scanning obtains can be detected and be identified to barrier, but due to cloud
Data scale is huge, can increase complexity to subsequent algorithm, so needing first to handle point cloud data.Laser radar data
Cluster is the basis of its point cloud data subsequent processing, at present the clustering method of laser radar have cluster based on K mean value, based on point
The cluster of layer, the cluster based on grid, closest cluster, cluster neural network based, is based on statistics at density clustering
Cluster and higher-dimension visualized data cluster etc..Density clustering method advantage is it can be found that arbitrary shape gathers
Class；The method of closest cluster calculates position selection that is simple, but depending greatly on first cluster centre；Based on net
Although the clustering method of lattice reduces calculation amount, real-time is improved, but away from the farther away barrier of laser radar, due to its cloud
Data reduce clustering precision than sparse.
China Patent No. CN201810047754.0 discloses a kind of barrier clustering method based on multi-line laser radar,
It is accurate to be improved in conjunction with density clustering algorithm DBSCAN using mesh filtering and adaptive neighborhood parameter for barrier clustering recognition
Property and real-time, but this method, when packing density is uneven, cluster pitch difference differs greatly, clustering result quality is poor.
China Patent No. CN201710318029.8 disclose the method for image zooming-out specification area a kind of, object identification and
Automobile is clustered using the algorithm of mesh connectivity zone marker, reduces calculation amount, improves real-time, but the party
Method is lower for remote data clusters precision, is also easy to produce more discrete point.
Summary of the invention
Goal of the invention: being directed to the deficiencies in the prior art, can be realized the present invention provides one kind to noise spot and
Sporadic data point is removed and simplifies the processing method of multiple barrier three-dimensional laser radar data of processing.
A kind of technical solution: processing method of multiple barrier three-dimensional laser radar data, comprising the following steps:
Step 1 establishes grating map: to establish grid coordinate with left and right directions in front of laser radar；It is a certain to choose laser radar
When the camera image of a frame point cloud data and synchronization angle inscribed；Above-mentioned point cloud data is projected in grid coordinate,
Whether after rejecting road surface point, point cloud data is projected to grid, count in each grid comprising point cloud projection；If there are points in grid
Then the grid is occupied state for cloud projection, indicates that there are barriers in the grid；Otherwise it is unoccupied state, indicates this grid not
There are barriers；The grid of occupied state, which is formed, constitutes original grating map, is checked with camera image；
Step 2, grid filtering: the hanging small barrier in environment sensing link is excluded；By laser radar point cloud data same
Difference in one grid between maximum height value and minimum height values is compared with the height threshold of setting, and difference is less than height
Threshold determination is small barrier, which is demarcated as unoccupied state；The grid of occupied state constitutes obvious barrier after filtering
Grating map, checked with camera image；
Step 3 determines adaptation threshold value: choosing another frame point cloud data and the moment with the same radar and angle in step 1
The camera image of angle projects in the grid coordinate of step 1 frame data, and carries out the filtering of step 2；Utilize the number
According to progress adaptive threshold cluster；By comparative test, the threshold value under different distance is selected to be clustered, by cluster result with
The image of synchronization compares, and it is the most close or equal to obtain clusters number and barrier actual quantity under different threshold values
Threshold value be adapt to threshold value；
Step 4, mesh connectivity labeled clusters: preliminary clusters are carried out to the grating map of step 2；Connected region is extended simultaneously
And carry out secondary cluster；It is checked with camera image.
Further, the unit size and grid of the grid coordinate established are determined according to the precision and range of realizing radar detection
Quantity, grid coordinate is, using horizontal plane as coordinate plane, it is identical to establish size using laser radar as origin in the step 1
Square grid coordinate, the size of single grid is the precision of data processing, and the size of grid and the number of grid determine number
According to range.
Further, according to realization required precision setting height threshold value in the step 2.The mesh installed according to laser radar
Vehicle passage capacity setting height threshold value, can further filter noise data and achieve the purpose that simplify data.
Further, the cluster of adaptive threshold described in the step 3 uses closest clustering algorithm, by first data
Initial point of the point as first object edge, as the first kind, since second data point, successively by each data point and
Its previous point is compared, and calculates the distance between two o'clock；If the distance of point-to-point transmission is less than or equal to the adaptation threshold in step 3
Value, then it is assumed that the point and former point belong to same class, which is added in current class；If the distance of point-to-point transmission is greater than in step 3
Adaptation threshold value, then it is assumed that the point is not belonging to current class, creates a new class, and using the point as the starting point of this new class,
All data points are according to said method judged, the frame data of acquisition are divided into several cluster point sets.Using most adjacent
Nearly clustering algorithm, which can be determined quickly, adapts to threshold value, provides foundation for the region extension in later step.
Further, in order to further also can preferably cluster the discrete part of distant place barrier together, front is obtained
Than more complete barrier profile and relative position information, the preliminary clusters in the step 4 are eight connectivity zone marker method；Institute
State connected region extension be eight connectivity region substantially, according to the adaptation threshold values of step 3, obtain grid threshold by formula scales
Value, and then the connected region after being expanded；If barrier grid corresponds to extended area range, there are remaining discrete barriers
Grid point then clusters discrete point and the grid for the same barrier.
The utility model has the advantages that processing method proposed by the present invention, realizes the denoising to noise spot and scattered point and simplifies processing；
Grid map under different distances can be clustered, barrier discrete part in distant place also can preferably cluster together, obtain
Vehicle front is than more complete barrier profile point and relative position information；Data fusion for laser radar and camera is spread
Pad, so that the automatic collision avoidance for Active collision avoidance system provides necessary foundation.
Detailed description of the invention
Fig. 1 is the flow chart of data processing of the invention；
Fig. 2 is the adaptive closest clustering algorithm flow chart of the present invention；
Fig. 3 is the schematic diagram of connected region of the present invention extension.
Specific embodiment
Describe embodiment in detail below with reference to accompanying drawings.
Shown in Fig. 1, a kind of processing method of multiple barrier three-dimensional laser radar data, comprising the following steps:
Step 1 establishes grating map: to establish grid coordinate with left and right directions in front of laser radar；It is a certain to choose laser radar
When the camera image of a frame point cloud data and synchronization angle inscribed；Above-mentioned point cloud data is projected in grid coordinate,
Whether after rejecting road surface point, point cloud data is projected to grid, count in each grid comprising point cloud projection；If there are points in grid
Then the grid is occupied state for cloud projection, indicates that there are barriers in the grid；Otherwise it is unoccupied state, indicates this grid not
There are barriers；The grid of occupied state, which is formed, constitutes original grating map, is checked with camera image；
Step 2, grid filtering: the hanging small barrier in environment sensing link is excluded；By laser radar point cloud data same
Difference in one grid between maximum height value and minimum height values is compared with the height threshold of setting, and difference is less than height
Threshold determination is small barrier, which is demarcated as unoccupied state；The grid of occupied state constitutes obvious barrier after filtering
Grating map, checked with camera image；
Step 3 determines adaptation threshold value: choosing another frame point cloud data and the moment with the same radar and angle in step 1
The camera image of angle projects in the grid coordinate of step 1 frame data, and carries out the filtering of step 2；Utilize the number
According to progress adaptive threshold cluster；By comparative test, the threshold value under different distance is selected to be clustered, by cluster result with
The image of synchronization compares, and it is the most close or equal to obtain clusters number and barrier actual quantity under different threshold values
Threshold value be adapt to threshold value；
Step 4, mesh connectivity labeled clusters: preliminary clusters are carried out to the grating map of step 2；Connected region is extended simultaneously
And carry out secondary cluster；It is checked with camera image.
The language environment that the present invention uses is python language.
The grating map that the present invention establishes is using laser radar as origin, and using the direction x as vehicle forward direction, the direction y is
Vehicle left, size are 400 × 400 lattice, and single grid size is 10cm × 10cm, i.e., comprising within vehicle front 40m and
Range within each 20m in left and right.
If this grid is set to 1 comprising point cloud projection in grid, indicate that there are barriers in grid；If not including in grid
Point cloud projection, this grid are set to 0, indicate that there is no barriers in this grid.
Algorithm flow are as follows: this grid (pm, pn) of if has a projection
GridMap=1
Wherein, GridMap indicates the value of grid
The unit size for the grid coordinate established and the quantity of grid are determined according to the precision and range of realizing radar detection, it is described
Grid coordinate is, using horizontal plane as coordinate plane, to establish the identical square grid of size using laser radar as origin in step 1
Coordinate, the size of single grid are the precision of data processing, the size of grid and the number determination data range of grid.
Maximum value and minimum value in grid be for laser radar point cloud data in this grid the maximum value of z coordinate and most
The difference of maxima and minima is compared by small value with threshold value, if being less than threshold value, is judged as hanging small barrier, by this
Grid is set to 0.
Algorithm flow are as follows:Max [pm, pn]-Min [pm, pn] > D1
GridMap=1
else
(pm, pn) is hanging small barrier, GridMap=0
Wherein, Max [pm, pn], Min [pm, pn] are respectively the maximum value and minimum value of the point cloud level degree in grid (pm, pn),
D1 indicates threshold value.
According to realization required precision setting height threshold value in the step 2.The purpose vehicle installed according to laser radar
Passage capacity setting height threshold value, can further filter noise data and achieve the purpose that simplify data.
As shown in Fig. 2, being clustered using the point cloud data of the second frame by adaptive k-nearest neighbor, constantly adjust
Whole threshold value D2, the image of cluster result and synchronization is compared, and obtains the adaptation that Clustering Effect is best under different distance
Threshold value DT；
The image with the same frame of laser radar data is obtained using CCD industrial camera, using adaptive thresholding algorithm to laser thunder
It is clustered up to point cloud data, cluster result is compared with image, constantly debugging is to obtain point cloud data under different distance
Threshold value double constant k, to obtain the threshold value D of point cloud dataTWith the cluster threshold value N of grating mapT。
The threshold value D of point cloud dataT=k*R, R are distance of the data point to laser radar, the cluster threshold value N of grating mapT=
DT/G。
Wherein, G is lattice dimensions size.
Wherein, closest clustering algorithm is using first data point as the initial point of first object edge, as
Each data point is successively compared by one kind since second data point with its previous point, calculate two o'clock between away from
From:
Wherein,Respectively indicate two neighbouring points.
If the distance of point-to-point transmission is less than or equal to threshold value, i.e.,, then it is assumed that the point and former point belong to same class, will
The point is added in current class；If the distance of point-to-point transmission is greater than threshold value, i.e.,, then it is assumed that the point is not belonging to current class, creation
One new class, and using the point as the starting point of this new class, all data points are according to said method judged, by acquisition
The frame data are divided into several cluster point sets.
Adaptive threshold described in the step 3 cluster use closest clustering algorithm, using first data point as
The initial point of first object edge, it is successively that each data point is previous with it since second data point as the first kind
A point is compared, and calculates the distance between two o'clock；If the distance of point-to-point transmission is less than or equal to the adaptation threshold value in step 3, recognize
Belong to same class for the point and former point, which is added in current class；If the distance of point-to-point transmission is greater than the adaptation threshold in step 3
Value, then it is assumed that the point is not belonging to current class, a new class is created, and using the point as the starting point of this new class, to all
Data point is according to said method judged, the frame data of acquisition are divided into several cluster point sets.It is calculated using closest cluster
Method, which can be determined quickly, adapts to threshold value, provides foundation for the region extension in later step.
The grid calculated under first frame grid map different distance clusters threshold value, whereinFor lattice dimensions.
On the basis of the eight connectivity zone marker of the first frame data, the grid that all grid point values are 1 is traversed, will tentatively be gathered
The eight connectivity region of class extends to, by this barrier grid correspond within the scope of extended area remaining from
Scattered barrier grid point cluster is the same barrier, discrete obstacle object point effectively can be classified as same class in this way.
Eight connectivity zone marker refers to upper and lower, left and right, upper left, upper right, lower-left, the bottom right of corresponding position, is adjacent position
It sets and oblique adjacent position, totally 8 directions.As shown in figure 3, eight connectivity region is extended to 5 × 5 regions: traversal all values are 1
Grid, if in its 5 × 5 region, have other Discrete Grids value be 1, then these grids are classified as one kind.
In order to further also can preferably cluster the discrete part of distant place barrier together, front is obtained than more complete
Barrier profile and relative position information, the preliminary clusters in the step 4 are eight connectivity zone marker method；The connected region
Domain extension be eight connectivity region substantially, according to the adaptation threshold values of step 3, obtain grid threshold value by formula scales, in turn
Connected region after being expanded；If barrier grid corresponds to extended area range there are remaining discrete barrier grid point,
Then discrete point and the grid are clustered as the same barrier.
Claims (5)
1. a kind of processing method of multiple barrier three-dimensional laser radar data, which comprises the following steps:
Step 1 establishes grating map: to establish grid coordinate with left and right directions in front of laser radar；It is a certain to choose laser radar
When the camera image of a frame point cloud data and synchronization angle inscribed；Above-mentioned point cloud data is projected in grid coordinate,
Whether after rejecting road surface point, point cloud data is projected to grid, count in each grid comprising point cloud projection；If there are points in grid
Then the grid is occupied state for cloud projection, indicates that there are barriers in the grid；Otherwise it is unoccupied state, indicates this grid not
There are barriers；The grid of occupied state, which is formed, constitutes original grating map, is checked with camera image；
Step 2, grid filtering: the hanging small barrier in environment sensing link is excluded；By laser radar point cloud data same
Difference in one grid between maximum height value and minimum height values is compared with the height threshold of setting, and difference is less than height
Threshold determination is small barrier, which is demarcated as unoccupied state；The grid of occupied state constitutes obvious barrier after filtering
Grating map, checked with camera image；
Step 3 determines adaptation threshold value: choosing another frame point cloud data and the moment with the same radar and angle in step 1
The camera image of angle projects in the grid coordinate of step 1 frame data, and carries out the filtering of step 2；Utilize the number
According to progress adaptive threshold cluster；By comparative test, the threshold value under different distance is selected to be clustered, by cluster result with
The image of synchronization compares, and it is the most close or equal to obtain clusters number and barrier actual quantity under different threshold values
Threshold value be adapt to threshold value；
Step 4, mesh connectivity labeled clusters: preliminary clusters are carried out to the grating map of step 2；Connected region is extended simultaneously
And carry out secondary cluster；It is checked with camera image.
2. the processing method of multiple barrier three-dimensional laser radar data according to claim 1, it is characterised in that: described
Grid coordinate is, using horizontal plane as coordinate plane, to establish the identical square grid of size using laser radar as origin in step 1
Coordinate, the size of single grid are the precision of data processing, the size of grid and the number determination data range of grid.
3. the processing method of multiple barrier three-dimensional laser radar data according to claim 1, it is characterised in that: described
According to realization required precision setting height threshold value in step 2.
4. the processing method of multiple barrier three-dimensional laser radar data according to claim 1, it is characterised in that: described
The cluster of adaptive threshold described in step 3 uses closest clustering algorithm, using first data point as first object edge
The initial point of edge since second data point, successively compares each data point and its previous point as the first kind
Compared with the distance between calculating two o'clock；If the distance of point-to-point transmission is less than or equal to the adaptation threshold value in step 3, then it is assumed that the point is with before
Any belongs to same class, which is added in current class；If the distance of point-to-point transmission is greater than the adaptation threshold value in step 3, then it is assumed that
The point is not belonging to current class, creates a new class, and using the point as the starting point of this new class, presses this to all data points
Method is judged, the frame data of acquisition are divided into several cluster point sets.
5. the processing method of multiple barrier three-dimensional laser radar data according to claim 1, it is characterised in that: described
Preliminary clusters in step 4 are eight connectivity zone marker method；Connected region extension be eight connectivity region substantially, according to
The adaptation threshold values of step 3 obtains grid threshold value by formula scales, and then the connected region after being expanded；If barrier grid
Lattice correspond to extended area range, and there are remaining discrete barrier grid points, then cluster discrete point and the grid for the same barrier
Hinder object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910398906.6A CN110208819A (en) | 2019-05-14 | 2019-05-14 | A kind of processing method of multiple barrier three-dimensional laser radar data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910398906.6A CN110208819A (en) | 2019-05-14 | 2019-05-14 | A kind of processing method of multiple barrier three-dimensional laser radar data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110208819A true CN110208819A (en) | 2019-09-06 |
Family
ID=67785922
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910398906.6A Pending CN110208819A (en) | 2019-05-14 | 2019-05-14 | A kind of processing method of multiple barrier three-dimensional laser radar data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110208819A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826512A (en) * | 2019-11-12 | 2020-02-21 | 深圳创维数字技术有限公司 | Ground obstacle detection method, ground obstacle detection device, and computer-readable storage medium |
| CN110824495A (en) * | 2019-11-20 | 2020-02-21 | 中国人民解放军国防科技大学 | Laser radar-based drosophila visual inspired three-dimensional moving target detection method |
| CN110889362A (en) * | 2019-11-21 | 2020-03-17 | 大连理工大学 | Obstacle detection method using grid map height information |
| CN110989631A (en) * | 2019-12-30 | 2020-04-10 | 科沃斯机器人股份有限公司 | Self-moving robot control method, device, self-moving robot and storage medium |
| CN111273316A (en) * | 2020-02-18 | 2020-06-12 | 中国科学院合肥物质科学研究院 | Multi-laser radar multi-view object detection method based on profile expansion fusion |
| CN112634487A (en) * | 2019-09-24 | 2021-04-09 | 北京百度网讯科技有限公司 | Method and apparatus for outputting information |
| CN112666946A (en) * | 2020-12-21 | 2021-04-16 | 杭州萤石软件有限公司 | Method for improving cleaning efficiency and cleaning robot |
| CN112666535A (en) * | 2021-01-12 | 2021-04-16 | 重庆长安汽车股份有限公司 | Environment sensing method and system based on multi-radar data fusion |
| CN112740225A (en) * | 2020-09-30 | 2021-04-30 | 华为技术有限公司 | Method and device for determining road surface elements |
| CN112882058A (en) * | 2021-01-08 | 2021-06-01 | 中国石油大学(华东) | Shipborne laser radar obstacle detection method based on variable-size grid map |
| CN113340314A (en) * | 2021-06-01 | 2021-09-03 | 苏州天准科技股份有限公司 | Local cost map generation method, storage medium and intelligent unmanned inspection vehicle |
| WO2021217646A1 (en) * | 2020-04-30 | 2021-11-04 | 华为技术有限公司 | Method and device for detecting free space for vehicle |
| CN113657205A (en) * | 2021-07-29 | 2021-11-16 | 北京中科慧眼科技有限公司 | Obstacle detection method and system based on motion grid and intelligent terminal |
| CN113749585A (en) * | 2020-05-28 | 2021-12-07 | 宁波方太厨具有限公司 | Semantic-based self-adaptive sweeping method for sweeping robot |
| CN113805597A (en) * | 2021-09-28 | 2021-12-17 | 福州大学 | Obstacle self-protection artificial potential field method local path planning method based on particle swarm optimization |
| CN114387585A (en) * | 2022-03-22 | 2022-04-22 | 新石器慧通（北京）科技有限公司 | Obstacle detection method, detection device, and travel device |
| CN114384492A (en) * | 2022-03-24 | 2022-04-22 | 北京一径科技有限公司 | Point cloud processing method and device for laser radar and storage medium |
| CN116796210A (en) * | 2023-08-25 | 2023-09-22 | 山东莱恩光电科技股份有限公司 | Barrier detection method based on laser radar |
| CN117491983A (en) * | 2024-01-02 | 2024-02-02 | 上海几何伙伴智能驾驶有限公司 | Method for realizing passable region boundary acquisition and target relative position discrimination |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070219720A1 (en) * | 2006-03-16 | 2007-09-20 | The Gray Insurance Company | Navigation and control system for autonomous vehicles |
| CN107064955A (en) * | 2017-04-19 | 2017-08-18 | 北京汽车集团有限公司 | barrier clustering method and device |
| CN108256577A (en) * | 2018-01-18 | 2018-07-06 | 东南大学 | A kind of barrier clustering method based on multi-line laser radar |
| CN108983248A (en) * | 2018-06-26 | 2018-12-11 | 长安大学 | It is a kind of that vehicle localization method is joined based on the net of 3D laser radar and V2X |
| CN109031346A (en) * | 2018-07-09 | 2018-12-18 | 江苏大学 | A kind of periphery parking position aided detection method based on 3D laser radar |
-
2019
- 2019-05-14 CN CN201910398906.6A patent/CN110208819A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070219720A1 (en) * | 2006-03-16 | 2007-09-20 | The Gray Insurance Company | Navigation and control system for autonomous vehicles |
| CN107064955A (en) * | 2017-04-19 | 2017-08-18 | 北京汽车集团有限公司 | barrier clustering method and device |
| CN108256577A (en) * | 2018-01-18 | 2018-07-06 | 东南大学 | A kind of barrier clustering method based on multi-line laser radar |
| CN108983248A (en) * | 2018-06-26 | 2018-12-11 | 长安大学 | It is a kind of that vehicle localization method is joined based on the net of 3D laser radar and V2X |
| CN109031346A (en) * | 2018-07-09 | 2018-12-18 | 江苏大学 | A kind of periphery parking position aided detection method based on 3D laser radar |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112634487A (en) * | 2019-09-24 | 2021-04-09 | 北京百度网讯科技有限公司 | Method and apparatus for outputting information |
| CN110826512B (en) * | 2019-11-12 | 2022-03-08 | 深圳创维数字技术有限公司 | Ground obstacle detection method, ground obstacle detection device, and computer-readable storage medium |
| CN110826512A (en) * | 2019-11-12 | 2020-02-21 | 深圳创维数字技术有限公司 | Ground obstacle detection method, ground obstacle detection device, and computer-readable storage medium |
| CN110824495A (en) * | 2019-11-20 | 2020-02-21 | 中国人民解放军国防科技大学 | Laser radar-based drosophila visual inspired three-dimensional moving target detection method |
| CN110889362A (en) * | 2019-11-21 | 2020-03-17 | 大连理工大学 | Obstacle detection method using grid map height information |
| CN110989631A (en) * | 2019-12-30 | 2020-04-10 | 科沃斯机器人股份有限公司 | Self-moving robot control method, device, self-moving robot and storage medium |
| CN110989631B (en) * | 2019-12-30 | 2022-07-12 | 科沃斯机器人股份有限公司 | Self-moving robot control method, device, self-moving robot and storage medium |
| CN111273316A (en) * | 2020-02-18 | 2020-06-12 | 中国科学院合肥物质科学研究院 | Multi-laser radar multi-view object detection method based on profile expansion fusion |
| WO2021217646A1 (en) * | 2020-04-30 | 2021-11-04 | 华为技术有限公司 | Method and device for detecting free space for vehicle |
| CN113749585A (en) * | 2020-05-28 | 2021-12-07 | 宁波方太厨具有限公司 | Semantic-based self-adaptive sweeping method for sweeping robot |
| CN112740225A (en) * | 2020-09-30 | 2021-04-30 | 华为技术有限公司 | Method and device for determining road surface elements |
| CN112740225B (en) * | 2020-09-30 | 2022-05-13 | 华为技术有限公司 | Method and device for determining road surface elements |
| CN112666946B (en) * | 2020-12-21 | 2024-05-31 | 杭州萤石软件有限公司 | Method for improving cleaning efficiency and cleaning robot |
| CN112666946A (en) * | 2020-12-21 | 2021-04-16 | 杭州萤石软件有限公司 | Method for improving cleaning efficiency and cleaning robot |
| CN112882058B (en) * | 2021-01-08 | 2022-09-20 | 中国石油大学(华东) | Shipborne laser radar obstacle detection method based on variable-size grid map |
| CN112882058A (en) * | 2021-01-08 | 2021-06-01 | 中国石油大学(华东) | Shipborne laser radar obstacle detection method based on variable-size grid map |
| CN112666535A (en) * | 2021-01-12 | 2021-04-16 | 重庆长安汽车股份有限公司 | Environment sensing method and system based on multi-radar data fusion |
| CN113340314A (en) * | 2021-06-01 | 2021-09-03 | 苏州天准科技股份有限公司 | Local cost map generation method, storage medium and intelligent unmanned inspection vehicle |
| CN113657205A (en) * | 2021-07-29 | 2021-11-16 | 北京中科慧眼科技有限公司 | Obstacle detection method and system based on motion grid and intelligent terminal |
| CN113657205B (en) * | 2021-07-29 | 2024-04-19 | 北京中科慧眼科技有限公司 | Obstacle detection method and system based on motion grid and intelligent terminal |
| CN113805597A (en) * | 2021-09-28 | 2021-12-17 | 福州大学 | Obstacle self-protection artificial potential field method local path planning method based on particle swarm optimization |
| CN114387585A (en) * | 2022-03-22 | 2022-04-22 | 新石器慧通（北京）科技有限公司 | Obstacle detection method, detection device, and travel device |
| CN114387585B (en) * | 2022-03-22 | 2022-07-05 | 新石器慧通（北京）科技有限公司 | Obstacle detection method, detection device, and travel device |
| CN114384492A (en) * | 2022-03-24 | 2022-04-22 | 北京一径科技有限公司 | Point cloud processing method and device for laser radar and storage medium |
| CN116796210A (en) * | 2023-08-25 | 2023-09-22 | 山东莱恩光电科技股份有限公司 | Barrier detection method based on laser radar |
| CN116796210B (en) * | 2023-08-25 | 2023-11-28 | 山东莱恩光电科技股份有限公司 | Barrier detection method based on laser radar |
| CN117491983A (en) * | 2024-01-02 | 2024-02-02 | 上海几何伙伴智能驾驶有限公司 | Method for realizing passable region boundary acquisition and target relative position discrimination |
| CN117491983B (en) * | 2024-01-02 | 2024-03-08 | 上海几何伙伴智能驾驶有限公司 | Method for realizing passable region boundary acquisition and target relative position discrimination |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110208819A (en) | A kind of processing method of multiple barrier three-dimensional laser radar data | |
| CN109858460B (en) | Lane line detection method based on three-dimensional laser radar | |
| CN110780305B (en) | Track cone detection and target point tracking method based on multi-line laser radar | |
| CN109541634B (en) | Path planning method and device and mobile device | |
| CN110084272B (en) | Cluster map creation method and repositioning method based on cluster map and position descriptor matching | |
| CN102930509B (en) | Intelligent filtering method for airborne laser point cloud data | |
| CN111079611B (en) | Automatic extraction method for road surface and marking line thereof | |
| CN107356933B (en) | Unstructured road detection method based on four-line laser radar | |
| CN108256577B (en) | Obstacle clustering method based on multi-line laser radar | |
| CN108873013B (en) | Method for acquiring passable road area by adopting multi-line laser radar | |
| WO2022141911A1 (en) | Roadside sensing unit-based method for quick recognition of dynamic target point cloud and point cloud segmentation | |
| CN102564431B (en) | Multi-sensor-fusion-based unstructured environment understanding method | |
| CN101852609B (en) | Ground obstacle detection method based on binocular stereo vision of robot | |
| CN103778429B (en) | Automatic extraction method for road information in a kind of Vehicle-borne Laser Scanning point cloud | |
| CN102339019B (en) | Intelligent wheel chair obstacle avoidance method based on fuzzy neural network | |
| CN110320504A (en) | A kind of unstructured road detection method based on laser radar point cloud statistics geometrical model | |
| CN103745436B (en) | LiDAR point cloud data shape filtering method based on regional prediction | |
| CN103714538B (en) | road edge detection method, device and vehicle | |
| CN111291708A (en) | Transformer substation inspection robot obstacle detection and identification method integrated with depth camera | |
| CN106780524A (en) | A kind of three-dimensional point cloud road boundary extraction method | |
| CN104992072A (en) | Operation land parcel automatic identification and area statistics method based on spatial mesh division | |
| CN103206957B (en) | The lane detection and tracking method of vehicular autonomous navigation | |
| CN108088445A (en) | 3 d grid map path planning system and method based on octree representation | |
| CN105551082A (en) | Method and device of pavement identification on the basis of laser-point cloud | |
| CN109001757A (en) | A kind of parking space intelligent detection method based on 2D laser radar |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |