CN107316308A - A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm - Google Patents
A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm Download PDFInfo
- Publication number
- CN107316308A CN107316308A CN201710501988.3A CN201710501988A CN107316308A CN 107316308 A CN107316308 A CN 107316308A CN 201710501988 A CN201710501988 A CN 201710501988A CN 107316308 A CN107316308 A CN 107316308A
- Authority
- CN
- China
- Prior art keywords
- grid
- matrix
- distance
- idle
- map
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
Abstract
The invention provides the clean robot map dividing method based on improved multi-path spectral clustering algorithm, including：Input parameter；Call on distance transform algorithm computation grid map the distance of available free grid between any two, build distance matrix；Based on the distance matrix, corresponding similar matrix is built using gaussian kernel function, and according to similar matrix structure degree matrix；According to similar matrix and degree matrix, the Laplacian Matrix of normalized；Calculate the corresponding characteristic vector of k eigenvalue of maximum, construction feature matrix before Laplacian Matrix；Eigenmatrix is standardized and obtains eigenmatrixWith eigenmatrixEvery a line as k tie up sample, clustered using algorithm；If by the eigenmatrix of standardizationM row vectors be assigned in the n-th cluster, then just m-th of idle grid is assigned in n-th of subregion；Export the grating map split.The present invention takes into full account the influence of neighbouring grid, improves the adaptivity of algorithm.
Description
Technical field
The invention belongs to clean robot technical field, more particularly to it is a kind of based on the clear of improved multi-path spectral clustering algorithm
Clean robot map dividing method.
Background technology
In recent years, with the development of intellectualized technology, clean robot, such as sweeping robot have entered into huge numbers of families,
It can autonomously carry out floor cleaning work, and not collided with the barrier in environment, simple to operate, automate journey
Degree is high, practical, and people can be helped to have been liberated from lengthy and jumbled cleaning out.At present, clean robot is obtained
To being widely applied.
The premise that clean robot can efficiently complete clean up task is the map with a description surrounding environment.In ring
In terms of the map building of border, robot can pass through sensor and the existing SLAM of combination (Simultaneous Localization
And Mapping) technology, the information of environment is obtained by using the mode such as learning along side, and set up description ring based on this
The grating map of border feature.After map foundation, sweeping robot just can carry out cleaning.
With the raising that people are required clean robot operating efficiency, the ability to work of individual machine people is obviously not
The requirement of people can be met, particularly under some large scale indoor environments, such as market, airport Waiting Lounge, railway station.Cause
This, the research of multiple cleaning machine person cooperative works receives increasing attention.In multiple cleaning machine person cooperative works
In, in order to reasonably give each robot to distribute task, it is necessary to reasonably be divided environmental map before cleaning task is carried out
Cut.At present, related research is not carried out to the technology in the prior art.
The content of the invention
In view of this, it is an object of the invention to provide a kind of segmentation side suitable for indoor cleaning machine people's environmental map
Method, to help to solve the problem of multiple clean robot tasks are distributed under large scale environment.
To achieve these goals, the technical scheme that the present invention is provided is as follows：
A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm, it is characterised in that described
Method includes：
S10, input parameter；
S20, call on distance transform algorithm computation grid map the distance of available free grid between any two, and as
Fundamental construction distance matrix S；
S30, based on distance matrix S, build corresponding similar matrix W using gaussian kernel function, and according to similar matrix W
Structure degree matrix D；
S40, according to similar matrix W and degree matrix D, the Laplacian Matrix L of normalized；
S50, calculate Laplacian Matrix before the corresponding characteristic vector of k eigenvalue of maximum, and using these characteristic vectors as
One eigenmatrix U=[u of fundamental construction1…uk]；
S60, eigenmatrix is standardized and obtains eigenmatrix
S70, with eigenmatrixEvery a line as k tie up sample, clustered using algorithm；
S80, if by the eigenmatrix of standardizationM row vectors be assigned in the n-th cluster, then just by m
Individual idle grid is assigned in n-th of subregion, wherein, m, n is respectively the integer more than or equal to 1.
S90, exports the grating map with k sub-regions split.
It is preferred that, in S10, input parameter is specially that the number k and scale parameter σ, k of the subregion needed for input are
Integer more than 1.
It is preferred that, S21, on grating map, according to from left to right, order from top to bottom is found in idle condition
The available free grid of institute, is respectively labeled as idle grid 1, idle grid 2 ... ..., idle grid N, N is the integer more than or equal to 1；
S22, with idle grid 1 for target grid, idle grid around it assigns weights 1, then by grid that weights are 1
The idle grid of surrounding assigns weights 2, constantly repeats, and is up to the available free grid on whole grating map is all assigned to weights
Only, the weights on each idle grid represent that these weights the distance between with idle grid 1, are constituted distance matrix S's by it
The first row；Then with idle grid 2 for target grid, using with the idle identical method of grid 1, the of generation distance matrix S
Two rows；By that analogy, until using idle grid N as target grid, generating distance matrix S Nth row；So as to constitute N × N away from
From matrix S.
It is preferred that, the distance between each two free time grid need to be only calculated only once, the distance value of each grid and itself
For 0.
It is preferred that, following formula calculating is respectively adopted in similar matrix W and degree matrix D in S30：
Similar matrix W is built using formula one, wherein 1≤i≤N, 1≤j≤N, i, j are integer,
(Wij=exp (- d2(si,sj)/2σ2)) (formula one)
Using the structure degree matrix D of formula two.
D=diag (β1,β2,…,βN), wherein
It is preferred that, the algorithm in S70 is k-means algorithms.
It is preferred that, S90 is specially：The grating map for having split k sub-regions is exported to clean robot, is each clear
Clean robot distributes different subregions.
It is preferred that, before S10, grating map is created using SLAM algorithms, wherein, the cell on grating map has
Three kinds of states：Idle condition, unknown state, it is occupied state.
In summary, the invention has the advantages that：
1st, by the present invention in that with distance transform algorithm come the distance between computation grid map overhead free potential grid lattice, fully examining
Considered barrier to idle grid distance from influence, improve the adaptivity of algorithm.
2nd, map dividing method highly versatile provided by the present invention, not only can be with the environment of segmenting structure, in half hitch
Also preferable segmentation effect can be obtained in the environment of structure；Large scale environmental map can be not only handled, small-sized room can be also tackled
Interior environment.
3rd, the present invention introduces the thought of spectral clustering the segmentation of environmental map, is provided for map segmentation problem a kind of new
Resolving ideas.
4th, from the point of view of the segmentation result of the present invention, region on original grating map all by individually separated, it
Global map disposably can be divided into some sub- maps, while the similarity met in sub- map is maximum, each height
Similarity between figure is minimum.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be attached to what is used required in embodiment
Figure is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for this area
For those of ordinary skill, on the premise of not paying creative work, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 (a)-(d) is the schematic diagram in the present invention using distance transform algorithm calculating distance value process；Fig. 1 (e) is to make
With the distance matrix constructed by the distance value calculated；Fig. 1 (f) is the similar matrix created using gaussian kernel function；
Fig. 2 is the flow chart based on improved multi-path spectral clustering algorithm under the indoor environment that proposes in the present invention；
Fig. 3 (a) is the original grating map of an expression indoor environment；Fig. 3 (b) is represented according to side proposed by the invention
Method the map is split after result.
Embodiment
In order that those skilled in the art more fully understands the technical scheme in the present invention, below in conjunction with accompanying drawing, to this
Technical scheme in inventive embodiments is clearly and completely described.Obviously, described embodiment is only the portion of the present invention
Divide embodiment, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making wound
The every other embodiment obtained under the premise of the property made work, should all belong to the scope of protection of the invention.
In the present invention, indoor environment is that global context is represented in the form of grating map, described improved multichannel spectrum
Clustering algorithm is a kind of clustering method for carrying out distance between computation-free cell using distance transform algorithm, described map point
Segmentation method be in the form of topological diagram by the node that global context is divided into some sub-regions, topological diagram represent be one piece from
By region.
In the present invention, three kinds of states of cell on grating map, grating map are created first with SLAM algorithms：It is empty
Not busy state, unknown state, it is occupied state.
Shown in Figure 1, it illustrates the spacing using each idle grid on distance transform algorithm computation grid map
From process.As shown in Fig. 1 (a), on grating map, there are the cell of three kinds of colors, i.e. grid：White, grey, black,
They represent idle condition, unknown state, are occupied state respectively.It is now to carry out cluster segmentation to idle grid.Such as Fig. 1
(a) shown in, one has 17 idle grids, and first with idle grid 1 for target grid, the idle grid around it assigns weights
1, weights are then assigned into weights 2 for the idle grid around 1 grid, constantly repeated, until all on whole grating map
Idle grid is all by untill being assigned to weights, and the weights on each free time grid just represent it the distance between with idle grid 1, such as
Shown in Fig. 1 (b).These weights are constituted to the first row of distance matrix (leading diagonal is all 0 symmetrical matrix)；Then with the free time
Grid 2 is target point, is repeated the above steps, and generates a new distance map, shown in such as Fig. 1 (c), then constitutes these weights
Second row of distance matrix；By that analogy, completed until distance matrix is created, final institute is as shown in Fig. 1 (d).Need exist for note
Distance between meaning, each two free time grid need to be only calculated only once, while the distance value of each grid and itself is 0.
Fig. 1 (e) is using the distance matrix constructed by the distance value calculated.
Then, using gaussian kernel function (Wij=exp (- d2(si,sj)/2σ2)), i is unequal with j, just can create expression
The similar matrix of unit compartment similarity degree, Fig. 1 (f) is the similar matrix.In gaussian kernel function, d (si,sj) represent
It is distance between two sample points, we are used in the distance value calculated in previous step herein, are used as two cells
Between distance value.The similar matrix obtained according to calculating, just can be according to multi-path spectral clustering algorithm to the free time on grating map
Cell is clustered, and is finally completed the segmentation of whole grating map.
As shown in Fig. 2 the flow chart based on improved multi-path spectral clustering algorithm under the indoor environment proposed in the present invention：
S10, input parameter.Specially：Number k (that is, the number of cluster) and the yardstick ginseng of subregion needed for input
Number σ, k are the integer more than 1.
S20, call on distance transform algorithm computation grid map the distance of available free grid between any two, and as
Fundamental construction distance matrix S, distance matrix S are symmetrical matrix.Specially：
S21, on grating map, according to from left to right, order from top to bottom finds having time in idle condition
Free potential grid lattice, are respectively labeled as idle grid 1, idle grid 2 ... ..., idle grid N, N is the integer more than or equal to 1；
S22, with idle grid 1 for target grid, idle grid around it assigns weights 1, then by grid that weights are 1
The idle grid of surrounding assigns weights 2, constantly repeats, and is up to the available free grid on whole grating map is all assigned to weights
Only, the weights on each idle grid represent that these weights the distance between with idle grid 1, are constituted distance matrix S's by it
The first row；Then with idle grid 2 for target grid, using with the idle identical method of grid 1, the of generation distance matrix S
Two rows；By that analogy, until using idle grid N as target grid, generating distance matrix S Nth row；So as to constitute N × N away from
From matrix S.It is noted herein that, the distance between each two free time grid need to be only calculated only once, each grid and its sheet
The distance value of body is 0.
S30, based on distance matrix S, build corresponding similar matrix W using gaussian kernel function, and according to similar matrix W
Structure degree matrix D.Specially：
S31, similar matrix W is built using formula one, wherein 1≤i≤N, 1≤j≤N, i, j are integer,
(Wij=exp (- d2(si,sj)/2σ2)) (formula one)
S32, using the structure degree matrix D of formula two,
D=diag (β1,β2,…,βN), wherein
S40, according to similar matrix W and degree matrix D, the Laplacian Matrix L of normalized.
S50, calculate Laplacian Matrix before the corresponding characteristic vector of k eigenvalue of maximum, and using these characteristic vectors as
One eigenmatrix U=[u of fundamental construction1…uk]。
S60, eigenmatrix is standardized and obtains eigenmatrix
S70, with eigenmatrixEvery a line as k tie up sample, gathered using clustering algorithms such as k-means
Class.
S80, if by the eigenmatrix of standardizationM row vectors be assigned in the n-th cluster, then just by m
Individual idle grid is assigned in n-th of subregion.Wherein, m, n are respectively the integer more than or equal to 1.
S90, exports the grating map with k sub-regions split.Specifically, the grid that k sub-regions will have been split
Lattice map is exported to clean robot, is that each clean robot distributes different subregions.
Shown in Figure 3, in order to verify the segmentation effect of this method, Fig. 3 (a) is the original grid of an expression indoor environment
Lattice map.Now, clear area on grating map is split using partitioning algorithm proposed by the present invention, the number of subregion
Mesh is set to 5 in advance.Fig. 3 (b) is the result schematic diagram of segmentation, from the point of view of the result of segmentation, the region on original grating map
All by individually separated, global map disposably can be divided into some individual sub- maps by it, while meeting sub- map
Interior similarity is maximum, and the similarity between each sub- map is minimum, and this is also the technique effect that the present invention wants to realize.
It is appreciated that embodiment as described herein can be by hardware, software, firmware, middleware, microcode or its any combination
To realize.For hardware implementation mode, processing unit can be at one or more application specific integrated circuits (ASIC), data signal
Manage device (DSP), digital signal processing device (DSPD), PLD (PLD), field programmable gate array (FPGA),
Processor, controller, microcontroller, microprocessor, it is designed to perform other electronic units or its group of function described herein
Realized in closing., can be by it when with software, firmware, middleware or microcode, program code or code segment to realize embodiment
Be stored in the machine readable media of such as storage assembly.
It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie
In the case of without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, embodiment all should be regarded as exemplary, and be nonrestrictive, the scope of the present invention is by appended power
Profit is required rather than described above is limited, it is intended that all in the implication and scope of the equivalency of claim by falling
Change is included in the present invention.Any reference in claim should not be considered as to the claim involved by limitation.
Moreover, it will be appreciated that although the present specification is described in terms of embodiments, not each embodiment is only wrapped
Containing an independent technical scheme, this narrating mode of specification is only that for clarity, those skilled in the art should
Using specification as an entirety, the technical solutions in the various embodiments may also be suitably combined, forms those skilled in the art
It may be appreciated other embodiment.
Claims (8)
1. a kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm, it is characterised in that the side
Method includes：
S10, input parameter；
S20, call on distance transform algorithm computation grid map the distance of available free grid between any two, and based on this
Build distance matrix S；
S30, based on distance matrix S, build corresponding similar matrix W using gaussian kernel function, and build according to similar matrix W
Spend matrix D；
S40, according to similar matrix W and degree matrix D, the Laplacian Matrix L of normalized；
S50, calculates the corresponding characteristic vector of k eigenvalue of maximum before Laplacian Matrix, and based on these characteristic vectors
Build an eigenmatrix U=[u1L uk]；
S60, eigenmatrix is standardized and obtains eigenmatrix U%；
S70, using eigenmatrix U% every a line as a k dimension sample, is clustered using algorithm；
S80, if the m row vectors of the eigenmatrix U% by standardization are assigned in the n-th cluster, then just empty by m-th
Free potential grid lattice are assigned in n-th of subregion, wherein, m, n is respectively the integer more than or equal to 1.
S90, exports the grating map with k sub-regions split.
2. according to the method described in claim 1, it is characterised in that：
In S10, input parameter is specially that the number k and scale parameter σ, k of the subregion needed for input are whole more than 1
Number.
3. according to the method described in claim 1, it is characterised in that：S20 is specifically included：
S21, on grating map, according to from left to right, order from top to bottom finds the available free grid of institute in idle condition
Lattice, are respectively labeled as idle grid 1, idle grid 2 ... ..., idle grid N, N is the integer more than or equal to 1；
S22, with idle grid 1 for target grid, the idle grid around it assigns weights 1, around the grid for being then 1 by weights
Idle grid assign weights 2, constantly repeat, until on whole grating map available free grid all by untill being assigned to weights,
Weights on each free time grid represent that these weights the distance between with idle grid 1, are constituted the first of distance matrix S by it
OK；Then with idle grid 2 for target grid, using with the idle identical method of grid 1, generation distance matrix S the second row；
By that analogy, until using idle grid N as target grid, generating distance matrix S Nth row；So as to constitute N × N apart from square
Battle array S.
4. method according to claim 3, it is characterised in that：Distance between each two free time grid need to only be calculated one
Secondary, the distance value of each grid and itself is 0.
5. according to the method described in claim 1, it is characterised in that：Similar matrix W and degree matrix D in S30 are respectively adopted down
Formula is calculated：
Similar matrix W is built using formula one, wherein 1≤i≤N, 1≤j≤N, i, j are integer,
(Wij=exp (- d2(si,sj)/2σ2)) (formula one)
Using the structure degree matrix D of formula two.
D=diag (β1,β2,L,βN), wherein
6. according to the method described in claim 1, it is characterised in that：Algorithm in S70 is k-means algorithms.
7. according to the method described in claim 1, it is characterised in that：S90 is specially：
The grating map for having split k sub-regions is exported to clean robot, is that each clean robot distributes different sons
Region.
8. the method according to any one in claim 1-7, it is characterised in that：
Before S10, grating map is created using SLAM algorithms, wherein, the cell on grating map has three kinds of states：It is empty
Not busy state, unknown state, it is occupied state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710501988.3A CN107316308A (en) | 2017-06-27 | 2017-06-27 | A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710501988.3A CN107316308A (en) | 2017-06-27 | 2017-06-27 | A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107316308A true CN107316308A (en) | 2017-11-03 |
Family
ID=60179555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710501988.3A Pending CN107316308A (en) | 2017-06-27 | 2017-06-27 | A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107316308A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107817802A (en) * | 2017-11-09 | 2018-03-20 | 北京进化者机器人科技有限公司 | The construction method and device of hybrid double-layer map |
| CN107833230A (en) * | 2017-11-09 | 2018-03-23 | 北京进化者机器人科技有限公司 | The generation method and device of indoor environment map |
| CN109214428A (en) * | 2018-08-13 | 2019-01-15 | 平安科技（深圳）有限公司 | Image partition method, device, computer equipment and computer storage medium |
| CN109409403A (en) * | 2018-09-12 | 2019-03-01 | 太原理工大学 | Brain network clustering method based on local attribute and topological structure |
| CN109887297A (en) * | 2019-04-23 | 2019-06-14 | 太原理工大学 | The division methods of urban traffic control sub-district based on quick global K-means spectral clustering |
| CN110044639A (en) * | 2019-04-24 | 2019-07-23 | 陕西重型汽车有限公司 | A kind of commercial vehicle MD-VTD system that segments market based on real vehicle operation big data platform |
| CN110726409A (en) * | 2019-09-09 | 2020-01-24 | 杭州电子科技大学 | Map fusion method based on laser SLAM and visual SLAM |
| CN111178646A (en) * | 2018-10-23 | 2020-05-19 | 广达电脑股份有限公司 | Task area allocation method for a plurality of cleaning devices and system thereof |
| WO2021238115A1 (en) * | 2020-05-26 | 2021-12-02 | 珠海一微半导体股份有限公司 | Method for establishing map traversal block of global grid map, chip, and mobile robot |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853491A (en) * | 2010-04-30 | 2010-10-06 | 西安电子科技大学 | SAR (Synthetic Aperture Radar) image segmentation method based on parallel sparse spectral clustering |
| CN102855638A (en) * | 2012-08-13 | 2013-01-02 | 苏州大学 | Detection method for abnormal behavior of vehicle based on spectrum clustering |
| CN103268729A (en) * | 2013-05-22 | 2013-08-28 | 北京工业大学 | Mobile robot cascading type map creating method based on mixed characteristics |
| CN106840168A (en) * | 2017-03-16 | 2017-06-13 | 苏州大学 | Complete coverage path planning method under clean robot and its dynamic environment |
-
2017
- 2017-06-27 CN CN201710501988.3A patent/CN107316308A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853491A (en) * | 2010-04-30 | 2010-10-06 | 西安电子科技大学 | SAR (Synthetic Aperture Radar) image segmentation method based on parallel sparse spectral clustering |
| CN102855638A (en) * | 2012-08-13 | 2013-01-02 | 苏州大学 | Detection method for abnormal behavior of vehicle based on spectrum clustering |
| CN103268729A (en) * | 2013-05-22 | 2013-08-28 | 北京工业大学 | Mobile robot cascading type map creating method based on mixed characteristics |
| CN106840168A (en) * | 2017-03-16 | 2017-06-13 | 苏州大学 | Complete coverage path planning method under clean robot and its dynamic environment |
Non-Patent Citations (1)
| Title |
|---|
| MING LIU ET AL.: "Regional topological segmentation based on Mutual Information Graphs", 《2011 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107817802A (en) * | 2017-11-09 | 2018-03-20 | 北京进化者机器人科技有限公司 | The construction method and device of hybrid double-layer map |
| CN107833230A (en) * | 2017-11-09 | 2018-03-23 | 北京进化者机器人科技有限公司 | The generation method and device of indoor environment map |
| CN109214428A (en) * | 2018-08-13 | 2019-01-15 | 平安科技（深圳）有限公司 | Image partition method, device, computer equipment and computer storage medium |
| CN109214428B (en) * | 2018-08-13 | 2023-12-26 | 平安科技（深圳）有限公司 | Image segmentation method, device, computer equipment and computer storage medium |
| CN109409403B (en) * | 2018-09-12 | 2019-07-26 | 太原理工大学 | Brain network clustering method based on local attribute and topological structure |
| CN109409403A (en) * | 2018-09-12 | 2019-03-01 | 太原理工大学 | Brain network clustering method based on local attribute and topological structure |
| CN111178646A (en) * | 2018-10-23 | 2020-05-19 | 广达电脑股份有限公司 | Task area allocation method for a plurality of cleaning devices and system thereof |
| CN111178646B (en) * | 2018-10-23 | 2023-04-07 | 广达电脑股份有限公司 | Task area allocation method for a plurality of cleaning devices and system thereof |
| CN109887297A (en) * | 2019-04-23 | 2019-06-14 | 太原理工大学 | The division methods of urban traffic control sub-district based on quick global K-means spectral clustering |
| CN109887297B (en) * | 2019-04-23 | 2021-03-26 | 太原理工大学 | Method for dividing urban traffic control subareas based on rapid global K-means spectral clustering |
| CN110044639A (en) * | 2019-04-24 | 2019-07-23 | 陕西重型汽车有限公司 | A kind of commercial vehicle MD-VTD system that segments market based on real vehicle operation big data platform |
| CN110726409A (en) * | 2019-09-09 | 2020-01-24 | 杭州电子科技大学 | Map fusion method based on laser SLAM and visual SLAM |
| CN110726409B (en) * | 2019-09-09 | 2021-06-22 | 杭州电子科技大学 | Map fusion method based on laser SLAM and visual SLAM |
| WO2021238115A1 (en) * | 2020-05-26 | 2021-12-02 | 珠海一微半导体股份有限公司 | Method for establishing map traversal block of global grid map, chip, and mobile robot |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107316308A (en) | A kind of clean robot map dividing method based on improved multi-path spectral clustering algorithm | |
| Tejani et al. | Multiobjective adaptive symbiotic organisms search for truss optimization problems | |
| Ventura et al. | Grouping parts and tools in flexible manufacturing systems production planning | |
| CN109214599B (en) | Method for predicting link of complex network | |
| CN104077311B (en) | Vehicle location indexing means and device | |
| CN102932479B (en) | Virtual network mapping method for realizing topology awareness based on historical data | |
| CN107292341A (en) | Adaptive multi views clustering method based on paired collaboration regularization and NMF | |
| CN107682871B (en) | Wireless sensor network relay node deployment optimization method and wireless sensor network | |
| CN108337685B (en) | Wireless sensor network data fusion method based on sub-clustering DGM | |
| Halász et al. | Dynamic redistribution of a swarm of robots among multiple sites | |
| US20200372412A1 (en) | System and methods to share machine learning functionality between cloud and an iot network | |
| CN106462905A (en) | System, method and apparatuses for identifying load volatility of a power customer and a tangible computer readable medium | |
| DE102020106136A1 (en) | SYSTEM AND PROCEDURE FOR AUTOMATIC COMMISSIONING OF ONE OR MORE NETWORKS OF ELECTRONIC DEVICES | |
| Cui et al. | Learning global pairwise interactions with Bayesian neural networks | |
| CN104898089B (en) | A kind of passive type localization method based on spatial migration compressed sensing | |
| CN115563674A (en) | Initial floorplan generation method and apparatus, electronic device, and storage medium | |
| CN108446712A (en) | ODN nets Intelligent planning method, apparatus and system | |
| CN115223662A (en) | Data processing method, device, equipment and storage medium | |
| Siddique et al. | Hybrid harmony search algorithms | |
| CN113344470A (en) | Intelligent power supply system management method based on block chain | |
| CN116894979A (en) | Graph classification method based on supervised adaptive soft clustering hierarchical pooling operator | |
| Ismail et al. | River flow forecasting: a hybrid model of self organizing maps and least square support vector machine | |
| CN116523001A (en) | Method, device and computer equipment for constructing weak line identification model of power grid | |
| CN115841229A (en) | Intelligent scheduling method of internet vehicle connecting system | |
| CN107330205A (en) | Complicated passage formula Mobile partition system track layout optimization equipment |
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 |
Application publication date: 20171103 |
|
| RJ01 | Rejection of invention patent application after publication |