CN108960738A - A kind of laser radar data clustering method under warehouse aisles environment - Google Patents
A kind of laser radar data clustering method under warehouse aisles environment Download PDFInfo
- Publication number
- CN108960738A CN108960738A CN201810784967.1A CN201810784967A CN108960738A CN 108960738 A CN108960738 A CN 108960738A CN 201810784967 A CN201810784967 A CN 201810784967A CN 108960738 A CN108960738 A CN 108960738A
- Authority
- CN
- China
- Prior art keywords
- data point
- data
- point
- laser radar
- accessed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
Landscapes
- Business, Economics & Management (AREA)
- Economics (AREA)
- Engineering & Computer Science (AREA)
- Marketing (AREA)
- Quality & Reliability (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Accounting & Taxation (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The invention discloses the laser radar data clustering methods under a kind of warehouse aisles environment, comprising: 1. initialization cores points m and Size of Neighborhood σ;2. it is the state not accessed that measurement data points, which are concentrated all data point markers,;It is to have accessed by the status indication of the point 3. choosing a state in sequence is the data point p not accessed;4. whether calculating in the neighborhood σ of data point p comprising m core points, it is to go in next step, data point p is otherwise labeled as noise spot, goes to step 2;5. by twice for being dimensioned to radar dispersion models return value of threshold value σ;6. judging that next data point q is whether in the neighborhood of data point p in sequence, if not going to step 2 if, otherwise data point q is labeled as having accessed, if data point q is in the core point i.e. neighborhood of data q comprising m data point, then update threshold value, furthermore cluster C is added in data point q, goes to step 5.
Description
Technical field
The invention belongs to the environmental modeling field of mobile robot, the laser thunder under a kind of warehouse aisles environment is specifically disclosed
Up to data clustering method.
Background technique
The warehouse aisles environment that package is put in two sides is common scene in warehouse.In the present context, robot localization and
The basis of navigation is the environmental model for establishing channel scene, package as component part necessary in environmental model, model
Precision directly affects the accuracy of environmental model.And clustering algorithm is as the first step for extracting package model, the result of cluster
Directly affect the accuracy of model extraction.Therefore, carrying out accurate cluster to laser radar data under warehouse aisles environment has
Significance.
There is fixed drift angle between the measurement data points of laser radar, but the actual range between data point additionally depends on biography
The distance between sensor and barrier.Under warehouse aisles environment, laser radar to the different measurement angles of same barrier away from
Defection has different measurement distances, this also causes the degree of scatter of laser radar data point different.Same type of data point point
Scattered degree difference, which causes to cluster under warehouse aisles environment, to be become difficult to achieve.
In existing clustering algorithm research, it is most widely used with the Density Clustering that K mean cluster and DBSCAN are represented.Wherein
K mean cluster is most widely used, but K mean cluster algorithm has ignored the incidence relation between laser radar consecutive number strong point,
Furthermore algorithm needs the number of clusters of specified cluster in advance, and above-mentioned two feature causes K mean cluster well suited logical in warehouse
Road environment.DBCSCAN algorithm divides data point, the characteristics of due to warehouse aisles environment, this cluster side using fixed threshold
Formula is easy mistake and divides data point point.
In conclusion current clustering algorithm is due to the considerations of lacking to channel environment and laser radar feature, it is difficult to suitable
Data clusters for laser radar under channel environment.Therefore, the deficiency for current algorithm under warehouse aisles environment, having must
It is proposed a kind of laser radar data clustering algorithm for being used in warehouse aisles.
Summary of the invention
In view of this, to solve the above-mentioned problems, the present invention provides the laser radar data under a kind of warehouse aisles environment
Clustering method.This method is easily achieved, is adaptable, can stablize and accurately click through to laser radar data under channel environment
Row cluster.
To achieve the above object and other purposes, the laser radar data that the present invention is provided under a kind of warehouse aisles environment is poly-
Class method, comprising the following steps:
Step 1. initializes core points m and Size of Neighborhood σ;
Step 2. is by measurement data point set P={ p1,p2,...,pnBe ranked up to 180 degree according to 0 degree of measurement angle or
The ascending sequence of person's angle is ranked up, and it is the shape not accessed that measurement data points, which are concentrated all data point markers,
State;
It is the data point p not accessed that step 3. once chooses a state in sequence, is by the status indication of the point
Access, if it is the data point not accessed that state, which has been not present, in measurement data points concentration at this time, algorithm is terminated, and completes poly-
Class;
Whether step 4. calculates in the neighborhood σ of data point p comprising m core points, is to go in next step, otherwise will count
Strong point p is labeled as noise spot, goes to step 2;
Step 5. creates the cluster C of data point p, and data point p is added in cluster C, is calculated according to radar dispersion models current
Threshold size, and by twice for being dimensioned to radar dispersion models return value of threshold value σ, σ=2f (φ, θ);
Step 6. judges that next data point q is whether in the neighborhood of data point p in sequence, if not going to step if
2, otherwise by data point q labeled as having accessed, if data point q is in the core point i.e. neighborhood of data point q comprising m data
Point then updates threshold value, and new threshold size is twice, i.e. σ=2f (φ, θ) of laser radar dispersion models return value at q,
Furthermore cluster C is added in data q, goes to step 5.
Preferably,
φiFor the angle of laser radar the i-th beam light and horizontal direction, φi+1For laser radar i+1 beam light and water
Square to angle, θ be barrier and vertical direction angle, d be laser radar position horizontal line and barrier hand over
The distance of point.
By adopting the above-described technical solution, the present invention has the advantage that:
The with good grounds laser radar principle of this hair and channel environment feature, propose lidar measurement dispersion models;
The present invention is dynamically determined clustering threshold value according to measurement dispersion models, proposes the clustering method of adaptive threshold.
Clustering method proposed by the present invention has higher operational efficiency under warehouse aisles environment, and improves cluster
The accuracy of arithmetic result.
Other advantages, target and feature of the invention will be illustrated in the following description to a certain extent, and
And to a certain extent, based on will be apparent to those skilled in the art to investigating hereafter, Huo Zheke
To be instructed from the practice of the present invention.Target of the invention and other advantages can be realized by following specification and
It obtains.
Detailed description of the invention
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing to the present invention make into
The detailed description of one step:
Fig. 1 is dispersion models schematic diagram 1;
Fig. 2 is dispersion models schematic diagram 2;
Fig. 3 is flow chart of the invention.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.
It should be noted that the basic conception that only the invention is illustrated in a schematic way is illustrated provided in the present embodiment,
Then only shown in schema with it is of the invention in related component rather than component count, shape and size when according to actual implementation draw
System, when actual implementation kenel, quantity and the ratio of each component can arbitrarily change for one kind, and its assembly layout kenel can also
It can be increasingly complex.
The present invention provides the laser radar data clustering method under a kind of warehouse aisles environment, and specific summary of the invention is as follows:
Lidar measurement dispersion models are initially set up, as illustrated in fig. 1 and 2, XOY indicates current global coordinate system, in XOY
Under coordinate system, it is assumed that laser radar position is C, and barrier is AB and is θ, laser radar i-th with the angle of vertical direction
The angle of beam light and horizontal direction isThe angle of i+1 beam light and horizontal direction isLaser radar position
Horizontal line and barrier meet at point B, BC length and are denoted as d,
According to Fig. 1, current institute's ranging can be described as follows from CA and BA.
Distance of the adjacent laser beam on barrier, that is,
Divergence between laser ranging point, can be described as following formula.
Similarly, the dispersion models under scene as shown in Figure 2 can also
To be described, in summary two class scenes, radar dispersion models can be described as following formula:
Clustering method based on measurement divergence variable threshold value:
The data set for remembering the acquisition of laser radar is Z={ z1,z2,...,zn, each data correspond to one side of laser radar
To data, which indicates measurement distance in this direction.Range data is converted to the data point set under rectangular coordinate system,
It is denoted as P={ p1,p2,...,pn}.Constant m is set and is surrounding's minimal point of core point, while defining σ is each vertex neighborhood threshold
Value, which be dynamically determined according to dispersion models.As shown in figure 3, considering that the variable threshold value clustering method of dispersion models is specific
Steps are as follows:
It is default value that step 1., which initializes core points m and Size of Neighborhood σ,;
Step 2. is by measurement data point set P={ p1,p2,...,pnBe ranked up to 180 degree according to 0 degree of measurement angle or
The ascending sequence of person's angle is ranked up, while and being the state not accessed by all data point markers of concentration;
It is the data point p not accessed that step 3. once chooses a state in sequence, is by the status indication of the point
Access, if it is the data point not accessed that state, which has been not present, in point concentration at this time, algorithm is terminated, and completes cluster;
Whether it includes m point that step 4 calculates in the neighborhood σ of measurement data points p, is gone in next step, otherwise by data
Point p is labeled as noise spot, goes to step 2;
Step 5. creates the cluster C of measurement data points p, and measurement data points p is added in cluster C, according to measurement dispersion models
Present threshold value size is calculated, and by twice for being dimensioned to model return value of threshold value σ, is shown below:
σ=2f (φ, θ)
Step 6. judges that next data point q is whether in the neighborhood of data point p in sequence, if not going to step if
2, otherwise by data point q labeled as having accessed, if data point q is in the core point i.e. neighborhood of data point q comprising m data
Point then updates threshold value, and cluster C furthermore is added in data point q, goes to step 5.
Finally, it is stated that the above examples are only used to illustrate the technical scheme of the present invention and are not limiting, although referring to compared with
Good embodiment describes the invention in detail, those skilled in the art should understand that, it can be to skill of the invention
Art scheme is modified or replaced equivalently, and without departing from the objective and range of the technical program, should all be covered in the present invention
Protection scope in.
Claims (2)
1. the laser radar data clustering method under a kind of warehouse aisles environment, which comprises the following steps:
Step 1. initializes core points m and Size of Neighborhood σ;
Step 2. is by measurement data point set P={ p1,p2,...,pnBe ranked up to 180 degree or angle according to 0 degree of measurement angle
It spends ascending sequence to be ranked up, and it is the state not accessed that measurement data points, which are concentrated all data point markers,;
It is the data point p not accessed that step 3. once chooses a state in sequence, is to have accessed by the status indication of the point,
If it is the data point not accessed that state, which has been not present, in measurement data points concentration at this time, algorithm is terminated, and completes cluster;
Whether step 4. calculates comprising m core points in the neighborhood σ of data point p, is gone in next step, otherwise by data point
P is labeled as noise spot, goes to step 2;
Step 5. creates the cluster C of data point p, and data point p is added in cluster C, calculates present threshold value according to radar dispersion models
Size, and by twice for being dimensioned to radar dispersion models return value of threshold value σ, σ=2f (φ, θ);
Step 6. judges that next data point q is whether in the neighborhood of data point p in sequence, no if not going to step 2 if
Then by data point q labeled as having accessed, if it includes m data point in the core point i.e. neighborhood of data point q that data point q, which is,
Threshold value is updated, new threshold size is twice, i.e. σ=2f (φ, θ) of laser radar dispersion models return value at q, furthermore will
Cluster C is added in data point q, goes to step 5.
2. the laser radar data clustering method under a kind of warehouse aisles environment according to claim 1, which is characterized in that
φiFor the angle of laser radar the i-th beam light and horizontal direction, φi+1For laser radar i+1 beam light and horizontal direction
Angle, θ be barrier and vertical direction angle, d be laser radar position horizontal line and barrier intersection point away from
From.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810784967.1A CN108960738B (en) | 2018-07-17 | 2018-07-17 | Laser radar data clustering method under warehouse channel environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810784967.1A CN108960738B (en) | 2018-07-17 | 2018-07-17 | Laser radar data clustering method under warehouse channel environment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108960738A true CN108960738A (en) | 2018-12-07 |
CN108960738B CN108960738B (en) | 2022-02-15 |
Family
ID=64495659
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810784967.1A Active CN108960738B (en) | 2018-07-17 | 2018-07-17 | Laser radar data clustering method under warehouse channel environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108960738B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109948979A (en) * | 2019-03-14 | 2019-06-28 | 广州蓝胖子机器人有限公司 | A kind of method, equipment and the storage medium of inventory's detection |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105866790A (en) * | 2016-04-07 | 2016-08-17 | 重庆大学 | Laser radar barrier identification method and system taking laser emission intensity into consideration |
CN106919955A (en) * | 2017-03-07 | 2017-07-04 | 江苏大学 | A kind of two points of K mean algorithms based on density criteria for classifying |
-
2018
- 2018-07-17 CN CN201810784967.1A patent/CN108960738B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105866790A (en) * | 2016-04-07 | 2016-08-17 | 重庆大学 | Laser radar barrier identification method and system taking laser emission intensity into consideration |
CN106919955A (en) * | 2017-03-07 | 2017-07-04 | 江苏大学 | A kind of two points of K mean algorithms based on density criteria for classifying |
Non-Patent Citations (3)
Title |
---|
SUN DIHUA等: "Adaptive KLD sampling based Monte Carlo localization", 《 2018 CHINESE CONTROL AND DECISION CONFERENCE (CCDC)》 * |
包瑞胜等: "基于优化DBSCAN算法的智能车载激光雷达数据处理技术研究", 《中国交通信息化》 * |
段建民等: "基于四线激光雷达的道路信息提取与目标检测", 《激光杂志》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109948979A (en) * | 2019-03-14 | 2019-06-28 | 广州蓝胖子机器人有限公司 | A kind of method, equipment and the storage medium of inventory's detection |
CN109948979B (en) * | 2019-03-14 | 2021-05-11 | 广州蓝胖子机器人有限公司 | Inventory detection method, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN108960738B (en) | 2022-02-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108958282B (en) | Three-dimensional space path planning method based on dynamic spherical window | |
CN106054900B (en) | The interim barrier-avoiding method of robot based on depth camera | |
CN103984037B (en) | The mobile robot obstacle detection method and device of view-based access control model | |
CN106525000B (en) | Roadmarking automation extracting method based on laser scanning discrete point intensity gradient | |
CN111539994B (en) | Particle filter repositioning method based on semantic likelihood estimation | |
CN109724603A (en) | A kind of Indoor Robot air navigation aid based on environmental characteristic detection | |
JP2019120927A (en) | Method and device for creating grid map | |
CN110073362A (en) | System and method for lane markings detection | |
CN102901500A (en) | Aircraft optimal path determination method based on mixed probability A star and agent | |
CN111880191B (en) | Map generation method based on multi-agent laser radar and visual information fusion | |
CN112346463B (en) | Unmanned vehicle path planning method based on speed sampling | |
CN103697895A (en) | Method for determining optimal path of flight vehicle based on self-adaptive A star algorithm | |
CN102324041B (en) | Pixel classification method, joint body gesture recognition method and mouse instruction generating method | |
JP6381137B2 (en) | Label detection apparatus, method, and program | |
CN110263713A (en) | Method for detecting lane lines, device, electronic equipment and storage medium | |
CN108470174A (en) | Method for obstacle segmentation and device, computer equipment and readable medium | |
CN105469445B (en) | A kind of step-length changeably drawing generating method | |
CN113924459B (en) | Dynamic sensor range detection for vehicle navigation | |
CN109001757A (en) | A kind of parking space intelligent detection method based on 2D laser radar | |
CN109960269A (en) | A kind of simplification environmental modeling method for serving unmanned plane independent navigation | |
Konrad et al. | Localization in digital maps for road course estimation using grid maps | |
Huber et al. | Fusion of LIDAR data and aerial imagery for automatic reconstruction of building surfaces | |
CN109215112A (en) | A kind of mask method of unilateral side point cloud model | |
CN106292656A (en) | A kind of environmental modeling method and device | |
CN105243354B (en) | A kind of vehicle checking method based on target feature point |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |