CN105424026B - A kind of indoor navigation localization method and system based on a cloud track - Google Patents
A kind of indoor navigation localization method and system based on a cloud track Download PDFInfo
- Publication number
- CN105424026B CN105424026B CN201510741200.7A CN201510741200A CN105424026B CN 105424026 B CN105424026 B CN 105424026B CN 201510741200 A CN201510741200 A CN 201510741200A CN 105424026 B CN105424026 B CN 105424026B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- track
- cloud
- point
- similitude
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Navigation (AREA)
Abstract
The invention discloses a kind of indoor navigation localization method based on a cloud track and system, wherein, the method includes:The capture point cloud track in motion process, and point cloud map is built according to multiple point cloud track;Destination is obtained, current location is positioned, and according to a cloud map planning guidance path;Whether detection actual motion shifts relative to guidance path, is to plan guidance path again.By the present invention in that building map with capture point cloud track during exercise, it is not necessary to positioned with indoor map by extra setting certain infrastructure, but and self dispose, position, strong robustness of to human and environment high with navigation accuracy.
Description
Technical field
The present invention relates to wireless communication field, especially, it is related to a kind of indoor navigation localization method based on a cloud track
With system.
Background technology
It is that indoor navigation services bright spot of greatest concern to find most short, most convenient the run trace for arriving at.Example
Such as, user can be potentially encountered following scene:" in meeting building, how I reach 2016 meeting rooms from current location" or:
" in this shopping center, how I reach the shop that I admires for a long time from my current position" however strong need
Ask, the design of pervasive accurately indoor navigation system is still extremely challenging technology with realization.
Indoor navigation system of the prior art is designed based on indoor locating system, is roughly divided into two classes:One class
Infrastructure, such as method based on received signals fingerprint, the method for visible light communication are to rely on, its application scenarios is limited
System;Another kind of is based on indoor map, such as inertial navigation method, computer vision methods.However, these systems are dependent on
In specific infrastructure and available indoor map.
The problem of infrastructure and indoor map is depended on for the work of indoor navigation system in the prior art, at present not yet
There is effective solution.
The content of the invention
The problem of infrastructure and indoor map is depended on for the work of indoor navigation system in the prior art, it is of the invention
Purpose is to propose a kind of indoor navigation localization method and system based on a cloud track, can be on foundation-free facility and indoor ground
Normal indoor navigation service is provided in the environment of figure.
Based on above-mentioned purpose, the technical scheme that the present invention is provided is as follows:
According to an aspect of the invention, there is provided a kind of indoor navigation localization method based on a cloud track.
Included according to a kind of indoor navigation localization method based on a cloud track that the present invention is provided:
The capture point cloud track in motion process, and point cloud map is built according to multiple point cloud track;
Destination is obtained, current location is positioned, and according to a cloud map planning guidance path;
Whether detection actual motion shifts relative to guidance path, is to plan guidance path again.
Wherein, the capture point cloud track in motion process, is to capture multiple images by assigned frequency in motion process, and
Point cloud track is extracted from multiple images.
Wherein, building point cloud map according to multiple point cloud track includes:
Choose two point cloud tracks, the similitude between two point cloud tracks of measurement;
Two point cloud tracks are merged according to similitude;
Continue other cloud tracks of selection to merge successively, build a point cloud map.
Also, the similitude between two point cloud tracks of measurement includes:
Turning point is positioned on a cloud track;
Every point cloud track is divided into by a plurality of cloud track line segment according to turning point;
Whether include common rail trace segments in two point cloud tracks of detection;
The similitude between two point cloud tracks is measured according to common rail trace segments.
Also, being merged two point cloud tracks according to similitude includes:
Compare length of the common rail trace segments on two point cloud track line segments of point cloud track;
When equal length of the common rail trace segments on the point cloud track line segments of two point cloud tracks, in two point cloud rails
Multiple point, multiple phases for being designated point on the point cloud track line segment of two point cloud tracks of contrast are specified in the same position of trace segments
Like property, and two point cloud tracks are merged according to similitude;
When length of the common rail trace segments on two point cloud track line segments of point cloud track is not waited, in point cloud more long
By the length interception sliding window of shorter point cloud track line segment on the line segment of track, and according to previous step in each sliding window
Suddenly the similitude of each point is contrasted, and is merged two point cloud tracks according to similitude.
Also, being merged two point cloud tracks according to similitude includes:
When similitude is higher than upper bound threshold value, judge that two point cloud track line segments are similar, carry out track merging;
When similitude is less than lower bound threshold value, judge that two point cloud track line segments are dissimilar, do not carry out track merging;
When similitude is between upper bound threshold value and lower bound threshold value, two similar feelings of point cloud track line segment are not judged
Condition, increase is designated multiple in the quantity of point and again the point cloud track line segments of two point cloud tracks of contrast and is designated the similar of point
Property.
Meanwhile, positioning current location is to position current location using trilateration, including:
Choose a cloud positioning current location;
Another cloud positioning current location is chosen, and current location is updated using gravity model appoach to two current locations;
Iteration carries out previous step, until the amendment to current location is less than error threshold.
Also, detection actual motion whether relative to guidance path shift including:
In actual motion current location is positioned by assigned frequency;
Judged whether to be located on guidance path according to current location, if otherwise next step;
Historical path according to actual motion judges whether to occur without matching, if otherwise judging actual motion relative to navigation
Path shifts.
In addition, also being informed with prompt message while planning guidance path again.
According to another aspect of the present invention, a kind of indoor navigation alignment system based on a cloud track is additionally provided.
It is as described above according to a kind of indoor navigation alignment system based on a cloud track that the present invention is provided.
From the above it can be seen that the technical scheme that the present invention is provided is by using capture point cloud track structure during exercise
Build map, it is not necessary to positioned with indoor map by extra setting certain infrastructure, but and self deployment, positioning
Strong robustness high with navigation accuracy, to human and environment.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to institute in embodiment
The accompanying drawing for needing to use is briefly described, it should be apparent that, drawings in the following description are only some implementations of the invention
Example, for those of ordinary skill in the art, on the premise of not paying creative work, can also obtain according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is a kind of flow chart of the indoor navigation localization method based on a cloud track according to the embodiment of the present invention;
Fig. 2 is a kind of structural relation of the indoor navigation localization method based on a cloud track according to the embodiment of the present invention
Figure;
During Fig. 3 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, different attitudes
Under capture point cloud and its number;
During Fig. 4 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, in office
With 375 Mass Distribution situation block diagrams of cloud are captured in market;
During Fig. 5 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, equipment attitude
With the correlation broken line graph of a cloud quality;
During Fig. 6 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, public sub- rail
Trace segments combination situation schematic diagram;
During Fig. 7 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, cloud computing is put
Relation block diagram between time and the degree of accuracy;
During Fig. 8 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, trilateration
Method and centralized positioning schematic diagram;
Fig. 9 is a kind of point cloud of the indoor navigation localization method based on a cloud track according to the embodiment of the present invention without matching
The broken line graph of skew;
During Figure 10 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, in office
With four test trails figures of the scene map in shopping center and each scene;
During Figure 11 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, scene map
In maximum tracking error and average tracking error statistical chart;
During Figure 12 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, scene map
Correlation block diagram between middle move distance and tracking error;
During Figure 13 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, scene map
Correlation block diagram between midpoint detection precision and point cloud;
During Figure 14 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, scene map
In whether there is collection point cloud quantity broken line graph under Attitude estimation;
During Figure 15 is a kind of indoor navigation localization method based on a cloud track according to the embodiment of the present invention, scene map
Point cloud quantity and the correlation block diagram for putting the cloud accuracy of map.
Specific embodiment
To make the object, technical solutions and advantages of the present invention become more apparent, below in conjunction with the embodiment of the present invention
Accompanying drawing, the technical scheme in the embodiment of the present invention is further carried out it is clear, complete, describe in detail, it is clear that it is described
Embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based on the embodiment in the present invention, this area
The every other embodiment that those of ordinary skill is obtained, belongs to the scope of protection of the invention.
The prosperity and development of mobile computing have promoted indoor navigation service as an attracting, promising application.Pass
The indoor navigation system of design of uniting is to rely on infrastructure, or is to rely on indoor map.The present invention is from five sides
Face has promoted indoor navigation system to move towards pervasive:Self can dispose, navigation accuracy high, the Shandong to environment dynamic change and personnel
Rod, without infrastructure and online start.The accurate positioning on a cloud map of point cloud localization method based on innovation is used
The position at family, the technical scheme user that can be navigated in the way of online startup reaches indoor impact point.Further
, additionally it is possible to run trace, run trace and cloud data combined structure point cloud map are extracted from cloud data.In addition, being
The system of raising goes to capture high-quality cloud to environmental dynamics and the robustness of personnel by the attitude for estimating equipment.
Under office and shopping mall scenario, the outstanding navigation performance of substantial amounts of experimental verification.
According to one embodiment of present invention, there is provided a kind of indoor navigation localization method based on a cloud track.
As shown in figure 1, the indoor navigation localization method based on a cloud track for providing according to embodiments of the present invention includes:
Step S101, the capture point cloud track in motion process, and point cloud map is built according to multiple point cloud track;
Step S103, obtains destination, positions current location, and according to a cloud map planning guidance path;
Whether step S105, detection actual motion shifts relative to guidance path, is to plan guidance path again.
Wherein, the capture point cloud track in motion process, is to capture multiple images by assigned frequency in motion process, and
Point cloud track is extracted from multiple images.
Wherein, building point cloud map according to multiple point cloud track includes:
Choose two point cloud tracks, the similitude between two point cloud tracks of measurement;
Two point cloud tracks are merged according to similitude;
Continue other cloud tracks of selection to merge successively, build a point cloud map.
Also, the similitude between two point cloud tracks of measurement includes:
Turning point is positioned on a cloud track;
Every point cloud track is divided into by a plurality of cloud track line segment according to turning point;
Whether include common rail trace segments in two point cloud tracks of detection;
The similitude between two point cloud tracks is measured according to common rail trace segments.
Also, being merged two point cloud tracks according to similitude includes:
Compare length of the common rail trace segments on two point cloud track line segments of point cloud track;
When equal length of the common rail trace segments on the point cloud track line segments of two point cloud tracks, in two point cloud rails
Multiple point, multiple phases for being designated point on the point cloud track line segment of two point cloud tracks of contrast are specified in the same position of trace segments
Like property, and two point cloud tracks are merged according to similitude;
When length of the common rail trace segments on two point cloud track line segments of point cloud track is not waited, in point cloud more long
By the length interception sliding window of shorter point cloud track line segment on the line segment of track, and according to previous step in each sliding window
Suddenly the similitude of each point is contrasted, and is merged two point cloud tracks according to similitude.
Also, being merged two point cloud tracks according to similitude includes:
When similitude is higher than upper bound threshold value, judge that two point cloud track line segments are similar, carry out track merging;
When similitude is less than lower bound threshold value, judge that two point cloud track line segments are dissimilar, do not carry out track merging;
When similitude is between upper bound threshold value and lower bound threshold value, two similar feelings of point cloud track line segment are not judged
Condition, increase is designated multiple in the quantity of point and again the point cloud track line segments of two point cloud tracks of contrast and is designated the similar of point
Property.
Meanwhile, positioning current location is to position current location using trilateration, including:
Choose a cloud positioning current location;
Another cloud positioning current location is chosen, and current location is updated using gravity model appoach to two current locations;
Iteration carries out previous step, until the amendment to current location is less than error threshold.
Also, detection actual motion whether relative to guidance path shift including:
In actual motion current location is positioned by assigned frequency;
Judged whether to be located on guidance path according to current location, if otherwise next step;
Historical path according to actual motion judges whether to occur without matching, if otherwise judging actual motion relative to navigation
Path shifts.
In addition, also being informed with prompt message while planning guidance path again.
Technical scheme is expanded on further below according to specific embodiment.
Fig. 2 is illustrated that the structural relation of indoor navigation of the invention (being named as pcNavi systems).As shown in Fig. 2
Indoor navigation system is mainly made up of 3 parts:Mobile client, Cloud Server and navigation user.
The track data of the sensor record of mobile client their daily lifes in space indoors.Specifically,
PcNavi systems calculate the run trace of user using related point cloud sequence.Meanwhile, the walking that the system goes out a cloud computing
Track is associated with point cloud.Importantly, mobile client captures high-quality point by detecting the attitude of mobile device
Cloud, and reduce calculating cost.The track of produced point cloud, uploads to Cloud Server and is further processed.
Cloud Server includes point cloud map structuring module and navigation module.We by calculate point cloud between similarity come
Similarity between measurement track, point cloud map is built by merging a plurality of track.Once the navigation requests from user are received,
Cloud Server calculates position of the user on a cloud map first, and uses the position as the starting point of navigation path.By pre-
The destination of definition, high in the clouds planning guidance path.When user walks to destination, high in the clouds receives to come from mobile client in real time
The cloud data at end, carries out the tracking of navigation process, offset detection.When user deviates navigation path, system provides warning, and
Again a walking path is planned.
Unique task of navigation user is exactly to be input into destination to system, then navigates to purpose under the guide of system
Ground.
Mobile client captures multiple images in motion process by assigned frequency, and point cloud rail is extracted from multiple images
Mark.Because point cloud is captured during walking, the quality of resulting point cloud is different.For example, when equipment is to landing
The point cloud captured when plate or ceiling is second-rate, and the quality of the point cloud against entity capture is higher.Catching under different attitudes
The feature for obtaining a cloud is counted out as shown in Figure 3.General, we are using the characteristic point extracted come the quality of metric point cloud.Fig. 4 shows
What is gone out is that 375 Mass Distribution situations of cloud are captured in office and market, and market captures invocation point cloud more than office
Quality will height, unconspicuous cloud of characteristic point redundant character that abundant point cloud has.In addition, when equipment picture head is downward upwards
When, capture be ceiling and floor point cloud, comparatively feature is not obvious.What other angles were mostly captured is bright feature
Aobvious point cloud (in addition to pure opposite wall).Equipment attitude shown in Fig. 5 demonstrates this point with the correlation of point cloud quality.
Cloud is significantly put in order to capture feature, we can carry out a cloud and capture with frame per second high, unobvious feature is filtered
Point cloud.But in this way, just cause that energy consumption increases.We pass through the pass between the attitude of facilities for observation capture point cloud and point cloud quality
System, devises simple and effective high-quality point cloud catching method, while not increasing energy consumption.
Our defining point cloud capturing events are f, are a binary class problems, i.e. capture or not capture point cloud.
We are returned using logistic and equipment attitude are predicted.Here, six independents variable, i.e. travel distance (γ d), three are had
Individual attitude angle, Fibre Optical Sensor, proximity transducer, i.e. X={ a, p, r, sl, sp, Δ (d) }, parameter sets are defined as θ.
Logistic regression equations such as following formula:
F=θTX
Further, equation is rewritten as using Sigmoid functions:
Its cost function is:
Our target is to minimize cost function.We are solved using gradient descent method, are obtained:
Wherein,
If additionally, during walking, system is not captured and obtains a cloud within a step, will continue to send capture point
The request of cloud.
For the point cloud for having obtained, we are calculated run trace using existing algorithm from a cloud sequence, definition
It is r, r={ x, y, pc }.Similitude between our point of use cloud measurement tracks, is to divide track for trajectory using turning point
Section, and the similitude between track is measured on the basis of in-orbit trace segments.Key issue is a measurement for cloud similitude.We
Substantial amounts of experiment is carried out, has had been found that the similitude between two clouds is more than at least 80, you can think that two clouds are
Similar.
Point cloud map is the set of the position relationship between the point cloud of indoor location entity.We utilize a plurality of cloud track
Merge a cloud map, the principle of merging is to minimize the diversity between point cloud track.In order to merge two tracks, we are first
First calculate public sub-trajectory line segment.Fig. 6 is by public sub-trajectory line segmentWithBe relatively divided into two kinds of situations, other feelings
Condition can reduction be both of these case.
Situation 1:That is the Case A in Fig. 6.Simplest method is that the point cloud on two orbit segments is counted one by one
Similitude is calculated, but so process is time-consuming.We devise random point cloud number selection algorithm (RPM) reduction and calculate complicated
Property.Concretism is a certain proportion of cloud of random selection, is compared two-by-two.If the number of similar point cloud is more than upper
, it is believed that two strip track line segments are similar.How the number of similar point cloud is less than lower bound, it is believed that two track line segments not phase
Seemingly.If the number of similar point cloud is between bound, it is believed that two track line segments are uncertain mutually dissimilar, also need to increase point
Cloud ratio carries out extra calculating.In a worst case scenario, it is all of cloud all by calculating similitude.
Situation 2:That is the Case B in Fig. 6.The sub-line section path length that two users are walked is unequal, I
Choose short sub-line section track on the basis of, similitude is calculated by the way of sliding window, until finding similar trajectory
Untill section.
In order to determine the ratio value of selected point cloud number, We conducted substantial amounts of experiment.We define two measurements
Index:One is the degree of accuracy, and another is time ratio.Fig. 7 is illustrated that the relation between cloud computing time and the degree of accuracy, such as
Shown in Fig. 7, during by the degree of accuracy and time ratio on the basis of ratio value, it is chosen at optimal between [0.6,0.8].
Once sufficient amount of public sub-trajectory Line segment detection is out, we build point cloud map using RPM algorithms.Herein
During, for speed-up computation speed, we use VF2 algorithms.In order to the coordinate for changing different tracks is tied to unified coordinate system
Under, we use Bursa Wolf models.At the same time, in order to improve the performance of RPM algorithms, we are accelerated using ICP algorithm.
After completing point cloud map, we can position the position of user as starting point and according to the indoor mesh of navigation user
Ground calculate navigation path.If navigation path is not unique, pcNavi calculates an optimal trajectory.PcNavi also has offset detection
Function, when user offsets correct navigation path, system can provide prompting, while the planning of the current location based on user one
The new guidance path of bar.
Embodiments of the invention carry out the positioning of user using the method based on trilateration.Due to the complexity of indoor environment
Property, it may appear that point cloud is much like on multiple positions, and we are referred to as pinpoint problems.In order to solve this problem, most intuitively solve
Certainly method is increased to an acquisition for cloud quantity.Therefore we devise PCLA location algorithms.For the cloud data for giving, adopt
The position of user can be positioned with following Trilateration methods, as shown in figure 8, its customer location O is:
O=T-1B
Wherein,
Because the point cloud of capture is in camera coordinates system, we can directly carry out determining for user using the coordinate of point cloud
Position.Computing formula is:
N cloud data is given, our optimization aim is that the position calculated by PCLA algorithms is offset with physical location
Error is minimized.We define optimization aim:
S.t.ni=rij × nj, rij ∈ R, i ≠ j,
Optimization Steps are as shown in Figure 8:
Initial, choose the position that cloud positions user.
The position of user is positioned in the point cloud do not chosen, position of the gravity model appoach as end user is chosen.
Iteration previous step, until current gravity model appoach defined location is less than with a step gravity model appoach defined location before
Threshold value, terminates.
Wherein, threshold value we be set as 0.1 meter by experiment.Such Optimization Steps, we can quickly it is unique really
Determine the position on the point cloud map that user is currently located.
We follow the trail of the process of user's navigation using the determination of the existing real time position of a mysorethorn.For offset detection, we
Two kinds of situations are divided into process:One is point cloud is without skew caused by matching, the second is user's actual shifts.
Fig. 9 is illustrated that broken line graph of a cloud without matching skew.The historical track that we are walked using user, given threshold
Test point cloud whether error hiding.
Think that a cloud error hiding is caused if fruit dot cloud orients the position come more than certain threshold value (such as 3 steps are remote) inclined
Move.In this way, system will again ask cloud data, the secondary positioning of user is carried out.
If user actually occurs skew, pcNavi can remind user, and plan a new path.
Meanwhile, we calculate optimal path using Dijstra ' s algorithms.
The effect to sort method of the present invention is evaluated below.
Figure 10 is illustrated that four test trails that we choose respectively under office and two, shopping center scene.
In office block, four volunteers are required to be walked along tetra- lines of A, B, C, D, and each track is recorded
At least 20 wheels.The length of these tracks is 30 meters, 40 meters, 65 meters, 90 meters.In our experiment, all of volunteer can
Smoothly arrived at along optimal track under pcNavi guidances.As shown in Figure 11 (a), the maximum tracking error of pcNavi
It it is 1.41 meters, average tracking error is 1.35 meters.It is (general that the pcNavi system deviations reach less tracking error less than 2 steps
It is the steps of 0.75m mono-) because compared to two dimensional image, the change of pcNavi point cloud environments and crowd are more sensitive.
At the mall, the tracks of four walkings are respectively M1, M2, M3, M4, and length is respectively 30 meters, 60 meters, 90 meters,
120 meters, per pass track at least records 20 wheels.In our experiment, all of volunteer can successfully arrive at destination.As schemed
Shown in 11 (b), at 1.20 meters, average tracking error is 1.04 meters to the maximum tracking error of pcNavi experience.Tracking error is in purchase
Thing center is smaller because the point cloud in shopping center enriches than in office building, and the point cloud of office building is less.
Additionally, it will be seen that tracking error is increased slightly with the increase of walking distance, therefore further experiment
Relation between test tracking error and walking distance.Every volunteer is required to be walked in the shopping center of office building sum, and
The tracking error under different travel distances is recorded, walking distance is changed into 360 meters of often wheel experiments from 0, and experiment is repeated 10 times, its knot
Fruit is as shown in Figure 12 (a) and 12 (b).Experimental result shows that the navigation skew that pcNavi can be detected is 2.85 meters, while actual
Deviation is 2.80 meters.It was found that the increase of the not increased walking distance of tracking error.Because the tracking drift for causing can
Eliminated with being often positioned in a location position of cloud graphical user.
Therefore, the degree of accuracy of tracking error detection and deviation can meet indoor navigation system.In practice, it is high-precision
Degree navigates also not necessarily, because user easily can with the naked eye find destination.Additionally, we analyse in depth influence because
Element and separate-blas estimation tracking error.
The positioning precision of point cloud map is most important to tracking and separate-blas estimation.Our passes in office and market building
Key point accuracy of detection is estimated, and, from 1 to 6, the number of respective point is from 80 to 100 for the number of the point cloud on a position
Individual, Figure 13 (a) and Figure 13 (b) is experimental result.It was found that point accuracy of detection increases with the quantity of a cloud.Work as a little
When the number of cloud increases to 5, critical point detection precision nearly reaches 100%.In addition, when positioning precision layer point cloud number is 5, as
Reduce for the position error that a cloud quantity increases.It will be apparent that point cloud is that enough pcNavi are promptly and accurately positioned at a little on a small quantity
The position of cloud graphical user.This result also indicates that it is fairly accurate that the parameter of deduction is set to PCLA.
Influence of the characteristic point of capture point cloud to point cloud matching precision is a factor for restriction pcNavi navigation accuracys.Will
Hope person have collected 200 groups of cloud datas, respectively according to there is Attitude estimation to be processed with without Attitude estimation two ways, its result
As shown in figure 14.As can be seen that cloud characteristic point quantity is put under having Attitude estimation being consistently higher than without Attitude estimation.Therefore, attitude is estimated
Meter contributes positively to abundant function and catches point cloud.
The influence of the navigation accuracy of the point cloud map of the accuracy that we further assess.Four volunteers are required doing
Public building and shopping center consumption build a cloud map over three days, have collected nearly 3000 groups of cloud datas.We weigh one
The influence of the point cloud accuracy of map, using root-mean-square error (RMSE).The two-dimensional position coordinate SPC of a given n cloud, computing formula
For
Its result as shown in figure 15, the office of RMSEs point cloud maps very little (for example ,≤0.56m) and the situation in market.
It is obvious that with the increase of number of users, the precision for putting cloud map also increases therewith.
According to another embodiment of the invention, there is provided a kind of indoor navigation alignment system based on a cloud track.
The indoor navigation alignment system based on a cloud track for providing according to embodiments of the present invention is as described above.
In sum, by means of above-mentioned technical proposal of the invention, built by using capture point cloud track during exercise
Map, it is not necessary to positioned with indoor map by extra setting certain infrastructure, interior has been promoted in terms of five
Navigation system moves towards pervasive:Self can dispose, navigation accuracy high, the robustness to environment dynamic change and personnel, without base
Infrastructure and online startup.The accurate position that user is positioned on a cloud map of point cloud localization method based on innovation, be
The user that can be navigated in the way of online startup of uniting reaches indoor impact point;In addition, it is also possible to row is extracted from cloud data
Walk track, run trace and cloud data combined structure point cloud map;In order to improve system to environmental dynamics and personnel
Robustness, he goes to capture high-quality cloud by estimating the attitude of equipment;It is substantial amounts of under office and shopping mall scenario
Experimental verification our systems outstanding navigation performance.
Those of ordinary skill in the art should be understood:Specific embodiment of the invention is the foregoing is only, and
The limitation present invention, all any modification, equivalent substitution and improvements within the spirit and principles in the present invention, done etc. are not used in,
Should be included within protection scope of the present invention.
Claims (8)
1. a kind of indoor navigation localization method based on a cloud track, it is characterised in that including:
The capture point cloud track in motion process, and point cloud map is built according to multiple point cloud track, wherein, it is described according to multiple
Point cloud track builds point cloud map to be included:Choose two point cloud tracks, the similitude between two point cloud tracks of measurement;According to institute
Similitude is stated to merge two point cloud tracks;Continue other cloud tracks of selection to merge successively, build a point cloud map;Its
In, the similitude between two point cloud tracks of the measurement includes:Turning point is positioned on described cloud track;According to described turn
Every described cloud track is divided into a plurality of cloud track line segment by curved point;Whether detect in two point clouds track includes public affairs
Common rail trace segments;The similitude between two point cloud tracks is measured according to the common rail trace segments;
Destination is obtained, current location is positioned, and according to described cloud map planning guidance path;
Whether detection actual motion shifts relative to guidance path, is to plan guidance path again.
2. a kind of indoor navigation localization method based on a cloud track according to claim 1, it is characterised in that it is described
Capture point cloud track in motion process, is to capture multiple images by assigned frequency in motion process, and from the multiple image
It is middle to extract point cloud track.
3. a kind of indoor navigation localization method based on a cloud track according to claim 1, it is characterised in that according to institute
Stating similitude includes two point cloud tracks merging:
Compare length of the common rail trace segments on the point cloud track line segment of two point clouds track;
When equal length of the common rail trace segments on the point cloud track line segment of two point clouds track, described two
Multiple point is specified in the same position of bar point cloud track line segment, contrasts multiple on the point cloud track line segment of two point clouds track
The similitude of point is designated, and is merged two point cloud tracks according to the similitude;
When length of the common rail trace segments on the point cloud track line segment of two point clouds track is not waited, more long
By the length interception sliding window of shorter point cloud track line segment on the line segment of point cloud track, and according to upper in each sliding window
One step contrasts the similitude of each point, and is merged two point cloud tracks according to the similitude.
4. a kind of indoor navigation localization method based on a cloud track according to claim 3, it is characterised in that according to institute
Stating similitude includes two point cloud tracks merging:
When similitude is higher than upper bound threshold value, judge that two point cloud track line segments are similar, carry out track merging;
When similitude is less than lower bound threshold value, judge that two point cloud track line segments are dissimilar, do not carry out track merging;
When similitude is between upper bound threshold value and lower bound threshold value, two similar situations of point cloud track line segment are not judged, increase
Plus be designated the quantity of point and contrast again on the point cloud track line segment of two point clouds track it is multiple be designated it is similar
Property.
5. a kind of indoor navigation localization method based on a cloud track according to claim 1, it is characterised in that described fixed
Position current location is to position current location using trilateration, including:
Choose a cloud positioning current location;
Another cloud positioning current location is chosen, and current location is updated using gravity model appoach to described two current locations;
Iteration carries out previous step, until the amendment to current location is less than error threshold.
6. a kind of indoor navigation localization method based on a cloud track according to claim 5, it is characterised in that the inspection
Survey actual motion whether relative to guidance path shift including:
In actual motion current location is positioned by assigned frequency;
Judged whether to be located on guidance path according to the current location, if otherwise next step;
Historical path according to actual motion judges whether to occur without matching, if otherwise judging actual motion relative to guidance path
Shift.
7. a kind of indoor navigation localization method based on a cloud track according to claim 1, it is characterised in that advise again
Also informed with prompt message while drawing guidance path.
8. a kind of indoor navigation alignment system based on a cloud track, it is characterised in that used as any in claim 1-7
The indoor navigation localization method based on a cloud track described in one.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510741200.7A CN105424026B (en) | 2015-11-04 | 2015-11-04 | A kind of indoor navigation localization method and system based on a cloud track |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510741200.7A CN105424026B (en) | 2015-11-04 | 2015-11-04 | A kind of indoor navigation localization method and system based on a cloud track |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105424026A CN105424026A (en) | 2016-03-23 |
CN105424026B true CN105424026B (en) | 2017-07-07 |
Family
ID=55502400
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510741200.7A Active CN105424026B (en) | 2015-11-04 | 2015-11-04 | A kind of indoor navigation localization method and system based on a cloud track |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105424026B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111323035A (en) * | 2019-12-18 | 2020-06-23 | 北京嘀嘀无限科技发展有限公司 | Detection method and detection device for driving yaw and readable storage medium |
Families Citing this family (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105955267A (en) * | 2016-05-11 | 2016-09-21 | 上海慧流云计算科技有限公司 | Motion control method and motion control system |
US11017610B2 (en) | 2016-05-18 | 2021-05-25 | Google Llc | System and method for fault detection and recovery for concurrent odometry and mapping |
US10890600B2 (en) | 2016-05-18 | 2021-01-12 | Google Llc | Real-time visual-inertial motion tracking fault detection |
US10802147B2 (en) * | 2016-05-18 | 2020-10-13 | Google Llc | System and method for concurrent odometry and mapping |
CN106338292A (en) * | 2016-09-30 | 2017-01-18 | 百度在线网络技术(北京)有限公司 | Walking path processing method and device |
CN107091639B (en) * | 2017-05-12 | 2019-11-12 | 北京航空航天大学 | A kind of total trajectory length long based on adaptive average window determines method |
CN107204014B (en) * | 2017-05-24 | 2019-06-07 | 京东方科技集团股份有限公司 | Localization method, device and the intelligent terminal of intelligent terminal |
CN109147044A (en) * | 2017-09-19 | 2019-01-04 | 上海华测导航技术股份有限公司 | A kind of system that power transmission tower and wireline inspection are carried out based on point cloud data |
CN109146836A (en) * | 2017-09-19 | 2019-01-04 | 上海华测导航技术股份有限公司 | A kind of system and method carrying out power transmission tower and wireline inspection based on point cloud data |
CN109143256A (en) * | 2017-09-19 | 2019-01-04 | 上海华测导航技术股份有限公司 | A method of power transmission tower and wireline inspection are carried out based on point cloud data |
CN110515089B (en) * | 2018-05-21 | 2023-06-02 | 华创车电技术中心股份有限公司 | Driving auxiliary method based on optical radar |
CN109066857B (en) * | 2018-08-15 | 2021-12-24 | 重庆七腾科技有限公司 | Method for charging patrol robot and charger robot |
TWI679511B (en) * | 2018-08-22 | 2019-12-11 | 和碩聯合科技股份有限公司 | Method and system for planning trajectory |
CN116972880A (en) * | 2019-03-18 | 2023-10-31 | 深圳市速腾聚创科技有限公司 | Precision detection device of positioning algorithm |
US11514610B2 (en) * | 2019-08-14 | 2022-11-29 | Tencent America LLC | Method and apparatus for point cloud coding |
CN110780678A (en) * | 2019-10-29 | 2020-02-11 | 无锡汉咏科技股份有限公司 | Unmanned aerial vehicle visual navigation control method based on point cloud data |
CN111798517A (en) * | 2020-07-01 | 2020-10-20 | 小狗电器互联网科技(北京)股份有限公司 | Method and device for determining travel track, readable medium and electronic equipment |
CN112764074B (en) * | 2020-12-30 | 2024-01-12 | 深圳中科讯联科技股份有限公司 | Method and device for positioning navigation track and electronic equipment |
CN116534059B (en) * | 2023-07-04 | 2023-09-08 | 深圳海星智驾科技有限公司 | Adaptive perception path decision method, device, computer equipment and storage medium |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101526356A (en) * | 2008-03-07 | 2009-09-09 | 佛山市顺德区顺达电脑厂有限公司 | Method for programming pedestrian navigation path |
CN102306300B (en) * | 2011-08-25 | 2013-04-10 | 武汉光庭信息技术有限公司 | Curve model-based method for identifying road with lossy shape |
US9633564B2 (en) * | 2012-09-27 | 2017-04-25 | Google Inc. | Determining changes in a driving environment based on vehicle behavior |
CN103884330B (en) * | 2012-12-21 | 2016-08-10 | 联想(北京)有限公司 | Information processing method, mobile electronic equipment, guiding equipment and server |
CN104536445B (en) * | 2014-12-19 | 2018-07-03 | 深圳先进技术研究院 | Mobile navigation method and system |
-
2015
- 2015-11-04 CN CN201510741200.7A patent/CN105424026B/en active Active
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111323035A (en) * | 2019-12-18 | 2020-06-23 | 北京嘀嘀无限科技发展有限公司 | Detection method and detection device for driving yaw and readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN105424026A (en) | 2016-03-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105424026B (en) | A kind of indoor navigation localization method and system based on a cloud track | |
Renaudin et al. | Evaluating indoor positioning systems in a shopping mall: The lessons learned from the IPIN 2018 competition | |
Dong et al. | ViNav: A vision-based indoor navigation system for smartphones | |
Xu et al. | Indoor localization via multi-modal sensing on smartphones | |
US9292936B2 (en) | Method and apparatus for determining location | |
Potortì et al. | Off-line evaluation of indoor positioning systems in different scenarios: The experiences from IPIN 2020 competition | |
CN102576064B (en) | Method and apparatus for identification of points of interest within a predefined area | |
Hilsenbeck et al. | Graph-based data fusion of pedometer and WiFi measurements for mobile indoor positioning | |
EP2885609B1 (en) | Crowd-sourcing indoor locations | |
Alzantot et al. | Crowdinside: Automatic construction of indoor floorplans | |
TWI485421B (en) | Map matching device, system and method | |
US10375517B2 (en) | Crowd sourced pathway maps | |
CN106574975A (en) | Trajectory matching using peripheral signal | |
Li et al. | IndoorWaze: A crowdsourcing-based context-aware indoor navigation system | |
CN103827632B (en) | Position detection device and position detection method | |
Papaioannou et al. | Tracking people in highly dynamic industrial environments | |
Elhamshary et al. | JustWalk: A crowdsourcing approach for the automatic construction of indoor floorplans | |
Radaelli et al. | Using cameras to improve wi-fi based indoor positioning | |
Takafuji et al. | Indoor localization utilizing tracking scanners and motion sensors | |
Vy et al. | Pedestrian indoor localization and tracking using hybrid Wi-Fi/PDR for iPhones | |
Gao et al. | Indoor map construction via mobile crowdsensing | |
Matsubayashi et al. | A method for estimating walking speed by using magnetic signature to grasp people flow in indoor passages | |
Zhao et al. | A sensor-service collaboration approach for target tracking in wireless camera networks | |
Zhou et al. | Magspider: A localization-free approach for constructing global indoor map for navigation purpose | |
Menke et al. | Multi-modal indoor positioning of mobile devices |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |