CN110514199A - A kind of winding detection method and device of SLAM system - Google Patents
A kind of winding detection method and device of SLAM system Download PDFInfo
- Publication number
- CN110514199A CN110514199A CN201910804329.6A CN201910804329A CN110514199A CN 110514199 A CN110514199 A CN 110514199A CN 201910804329 A CN201910804329 A CN 201910804329A CN 110514199 A CN110514199 A CN 110514199A
- Authority
- CN
- China
- Prior art keywords
- carrier
- candidate
- winding
- location
- loopback location
- 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
- 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
- G01C21/04—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by terrestrial means
- G01C21/08—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by terrestrial means involving use of the magnetic field of the earth
-
- 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
- G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
- G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
- G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
- G01C21/165—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation combined with non-inertial navigation instruments
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geology (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
This application discloses a kind of winding detection method and device of SLAM system, it include: the environmental characteristic for obtaining carrier local environment and the candidate loopback location for obtaining the motion feature of the carrier, the carrier based on determined by bag of words, according to the environmental characteristic and motion feature, the winding detection confidence parameter for the candidate loopback location determined based on bag of words is calculated, winding detects the confidence level that the candidate loopback location of confidence parameter characterization is the history loopback location that carrier once passed through；If winding, which detects confidence parameter, is greater than preset threshold, which is determined as winding testing result.It can be seen that, according to the environmental characteristic and motion feature of carrier, it can determine that candidate loopback location is the confidence level for the history loopback location that carrier once passed through, in turn, after confidence level higher winding detection position is determined as final winding testing result, the accuracy for the winding testing result finally determined can be improved.
Description
Technical field
This application involves winding detection technique fields, more particularly to the winding detection method and dress of a kind of SLAM system
It sets.
Background technique
Currently, SLAM (Simultaneous Localization And Mapping, instant positioning and map structuring) system
System is commonly used in destination carrier in nothing or weak GPS (Global Positioning System, global positioning system) environment
Positioning and navigation.Wherein, destination carrier usually can be the autonomous such as unmanned vehicle, unmanned plane, unmanned boat or robot load
Body.And in SLAM system, global optimization precision and target location accuracy are usually improved using winding detection method.
Winding detection, and it is properly termed as closed loop detection, refer to that autonomous carrier recognition is presently in scene as once institute
The scene reached, so that the map that autonomous carrier is established in moving process forms closed loop.
In practical application, the winding detection of autonomous carrier can be carried out using bag of words method, being will be autonomous
The feature in image that mobile vehicle institute real-time perception obtains is classified as word, and by the corresponding word of the figure with pre-establish
Dictionary in word be compared, with determine autonomous carrier be currently located scene and the scene that once reached whether one
It causes, to realize that winding detects.But winding detection is carried out using bag of words method, the standard of acquired winding testing result
True property is not high, be easy to cause erroneous judgement.
Summary of the invention
The embodiment of the present application provides a kind of method and device of the winding detection of SLAM system, it is intended to improve and be based on bag of words
The accuracy of winding testing result determined by model.
In a first aspect, the embodiment of the present application provides a kind of winding detection method of SLAM system, which comprises
It obtains the environmental characteristic of carrier local environment and obtains the motion feature of the carrier, determined based on bag of words
The carrier candidate loopback location；
According to the environmental characteristic and the motion feature, the winding inspection for being directed to the candidate loopback location is calculated
Confidence parameter is surveyed, it is the history winding that the carrier passes through that the winding, which detects the confidence parameter characterization candidate loopback location,
The confidence level of position；
If the winding detection confidence parameter is greater than preset threshold, the candidate loopback location is determined as winding detection
As a result.
In some possible embodiments, the motion feature includes the pitch angle, roll angle and boat of the carrier
To angle, the environmental characteristic includes the Magnetic Field of the carrier local environment.
In some possible embodiments, the motion feature for obtaining carrier, comprising:
Obtain the ratio force information that the accelerometer being installed on the carrier is exported；
According to described than force information and the Magnetic Field, the pitch angle, roll angle and boat of the carrier are calculated
To angle.
In some possible embodiments, described according to the environmental characteristic and the motion feature, calculate needle
Confidence parameter is detected for the winding of the candidate loopback location, comprising:
According to the environmental characteristic and the motion feature, determine the carrier in the quantizating index of the current location
Vector；
Obtain quantizating index vector corresponding to the candidate loopback location；
According to the carrier in the quantizating index vector and the candidate loopback location of the current location corresponding to
The winding detection confidence parameter for being directed to the candidate loopback location is calculated in quantizating index vector.
In some possible embodiments, it is described according to the carrier the current location quantizating index vector with
And quantizating index vector corresponding to the candidate loopback location, the winding inspection for being directed to the candidate loopback location is calculated
Survey confidence parameter, comprising:
The carrier is calculated corresponding to the quantizating index vector and the candidate loopback location of the current location
The absolute value of the difference of the mould length of each component in quantizating index vector；
According to the absolute value of the difference of the mould length of each component, it is calculated and is directed to returning for the candidate loopback location
Ring detects confidence parameter.
Second aspect, the embodiment of the present application also provides a kind of winding detection device of SLAM system, described device includes:
Module is obtained, for obtaining the environmental characteristic of carrier local environment and obtaining motion feature, the base of the carrier
In the candidate loopback location for the carrier that bag of words determine；
Computing module is directed to described candidate time for calculating according to the environmental characteristic and the motion feature
The winding of ring position detects confidence parameter, and it is the carrier that the winding, which detects the confidence parameter characterization candidate loopback location,
The confidence level of the history loopback location of process；
Determining module, if being greater than preset threshold for winding detection confidence parameter, by the candidate loopback location
It is determined as winding testing result.
In some possible embodiments, the motion feature includes the pitch angle, roll angle and boat of the carrier
To angle, the environmental characteristic includes the Magnetic Field of the carrier local environment.
In some possible embodiments, the acquisition module, comprising:
First acquisition unit, the ratio force information exported for obtaining the accelerometer being installed on the carrier；
First computing unit, for than force information and the Magnetic Field, calculating bowing for the carrier according to described
The elevation angle, roll angle and course angle.
In some possible embodiments, the computing module, comprising:
Determination unit, for determining the carrier described current according to the environmental characteristic and the motion feature
The quantizating index vector of position；
Second acquisition unit, for obtaining quantizating index vector corresponding to the candidate loopback location；
Second computing unit, for according to the carrier in the current location quantizating index vector and the candidate
The winding detection confidence ginseng for being directed to the candidate loopback location is calculated in quantizating index vector corresponding to loopback location
Number.
In some possible embodiments, second computing unit, comprising:
First computation subunit, for calculating quantizating index vector and the time of the carrier in the current location
Select the absolute value of the difference of the mould length of each component in quantizating index vector corresponding to loopback location；
Second computation subunit is calculated and is directed to for the absolute value of the difference according to the mould length of each component
The winding of candidate's loopback location detects confidence parameter.
In the above-mentioned implementation of the embodiment of the present application, according to the environmental characteristic of carrier local environment and the carrier
Motion feature determines the confidence level of candidate's loopback location based on determined by bag of words, to improve according to the confidence level
The finally accuracy of determining winding testing result.Specifically, environmental characteristic and the acquisition of available carrier local environment
The candidate loopback location of the motion feature of the carrier, the carrier based on determined by bag of words, then, according to the environmental characteristic
And motion feature, the winding detection confidence for the candidate loopback location determined based on bag of words can be further calculated out
Parameter, wherein the winding detects the history winding position that confidence parameter characterization candidate's loopback location once passed through for the carrier
The confidence level set；If winding, which detects confidence parameter, is greater than preset threshold, which is determined as winding detection knot
Fruit.Although can make as it can be seen that the candidate loopback location for obtaining carrier based on bag of words is directly determined as winding testing result
Finally the accuracy of determining winding testing result is lower, but according to the fortune of the environmental characteristic of carrier local environment and carrier
Dynamic feature can determine that candidate's loopback location is the confidence level for the history loopback location that carrier once passed through.It is appreciated that
A possibility that confidence level is higher, shows the history loopback location that candidate's loopback location once passed through for carrier is higher, that is, carries
Body is presently in scene and is more likely to be the scene once passed through, thus using candidate's loopback location as winding testing result
Accuracy it is also higher, in turn, can after confidence level higher winding detection position is determined as final winding testing result
To improve the accuracy for the winding testing result finally determined.
Detailed description of the invention
In order to more clearly explain the technical solutions in the embodiments of the present application, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations as described in this application
Example, for those of ordinary skill in the art, is also possible to obtain other drawings based on these drawings.
Fig. 1 is an exemplary application schematic diagram of a scenario in the embodiment of the present application；
Fig. 2 is a kind of flow diagram of the winding detection method of SLAM system in the embodiment of the present application；
Fig. 3 is a kind of structural schematic diagram of the winding detection device of SLAM system in the embodiment of the present application.
Specific embodiment
In practical application, when carrying out winding detection based on bag of words come the carrier to autonomous, it is easy to can go out
The testing result of existing false positive or false negative.Wherein, false positive, refer to mistake by carrier without scene (be also
The scene that carrier is presently in) it is identified as the scene once passed through；And false negative, refer to the unidentified field that carrier iterates through out
Scape, i.e., it is unidentified go out the scene that is presently in of carrier be scene that the carrier once passed through, alternatively, mistake by the current institute of carrier
The scene at place is mapped with some practical different scene in the past.Therefore, it is based on existing bag of words obtained time
Ring testing result, accuracy be not high.
Based on this, the embodiment of the present application provides a kind of winding detection method of SLAM system, according to carrier local environment
Environmental characteristic and the carrier motion feature, determine candidate's loopback location based on determined by bag of words confidence level,
To improve the accuracy of finally determining winding testing result according to the confidence level.Specifically, locating for available carrier
The environmental characteristic of environment and the candidate winding position for obtaining the motion feature of the carrier, the carrier based on determined by bag of words
It sets, then, according to the environmental characteristic and motion feature, the candidate determined based on bag of words can be further calculated out
The winding of loopback location detects confidence parameter, wherein the winding detects confidence parameter characterization candidate's loopback location as the load
The confidence level for the history loopback location that body once passed through；If winding, which detects confidence parameter, is greater than preset threshold, which is returned
Ring position is determined as winding testing result.Although as it can be seen that directly that the candidate loopback location for obtaining carrier based on bag of words is true
It is set to winding testing result, the accuracy of the winding testing result finally determined can be made lower, but the ring according to locating for carrier
The environmental characteristic in border and the motion feature of carrier can determine that candidate's loopback location is that the history that carrier once passed through is returned
The confidence level of ring position.It is appreciated that the confidence level is higher, show that candidate's loopback location is that the history that carrier once passed through is returned
A possibility that ring position, is higher, i.e., carrier is presently in scene and is more likely to be the scene once passed through, thus with the candidate
Loopback location is also higher as the accuracy of winding testing result, and in turn, confidence level higher winding detection position is determined
After final winding testing result, the accuracy for the winding testing result finally determined can be improved.
As an example, the embodiment of the present application can be applied to exemplary application scene as shown in Figure 1.In the scene
In, user 101 can indicate that (such as can use remote controler or other controllers indicate etc.) robot 102 moves
It is dynamic to build figure；Since robot 102 can have some cumulative errors during building figure, robot 102, which can use, to be based on
The winding of bag of words detects to optimize pose.Specifically, the robot 102 that robot 102 can be determined based on bag of words
Candidate loopback location, since candidate's loopback location might not be accurate, if directly using candidate's loopback location as final
Winding testing result, a possibility that false positive or false negative occur, is higher, and therefore, robot 102 can also obtain currently
The environmental characteristic of itself local environment and the motion feature for obtaining itself, and according to the environmental characteristic and motion feature, meter
The winding detection confidence parameter for being directed to candidate loopback location is calculated, winding detection detects the candidate winding of confidence parameter characterization
Position is the confidence level for the history loopback location that robot 102 once passed through；If the winding, which detects confidence parameter, is greater than default threshold
Value, then candidate's loopback location can be determined as final winding testing result by robot 102.
It is understood that above-mentioned scene is only a Sample Scenario provided by the embodiments of the present application, the embodiment of the present application
It is not limited to this scene.For example, the technical solution of the embodiment of the present application is also possible to apply in other possible application scenarios
It can be in the carrier of autonomous in unmanned plane, unmanned vehicle, unmanned boat etc..For another example, in other possible application scenarios,
Can be is that corresponding operation interface is provided in robot, is realized directly in the operation interface to machine by user 101
The control etc. of people.In short, the embodiment of the present application can be applied in any scene applicatory, and it is not limited to above-mentioned scene and shows
Example.
In order to make the above objects, features, and advantages of the present application more apparent, below in conjunction with attached drawing to this Shen
Please the various non-limiting implementations in embodiment illustrate.Obviously, described embodiment is the application one
Section Example, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not doing
All other embodiment obtained under the premise of creative work out, shall fall in the protection scope of this application.
Referring to Fig.2, Fig. 2 shows the signals of the process of the winding detection method of SLAM system a kind of in the embodiment of the present application
Figure, this method can specifically include:
S201: obtaining the environmental characteristic of carrier local environment and obtains the motion feature of the carrier, based on bag of words
The candidate loopback location of the determining carrier.
Due in practical application, if directly the candidate loopback location determined based on bag of words is determined as final
Winding testing result, it is easy to false positive or false negative occur, and make the accuracy of winding testing result lower.For this purpose,
In the present embodiment, confidence level is based on to the candidate loopback location determined based on bag of words and is screened, to improve winding
The accuracy of testing result.It should be noted that the candidate loopback location for determining carrier based on bag of words, existing skill
There is corresponding specific implementation process in art, details are not described herein.Also, carrier as described in this embodiment, specifically can be
Carrier with autonomous ability, such as unmanned plane, unmanned vehicle, unmanned boat and robot.
In the present embodiment, the motion feature of environmental characteristic and the carrier based on carrier current environment can be,
The confidence level of candidate loopback location is determined, to screen to the candidate loopback location for being suitable as winding testing result.
Therefore, while obtaining candidate loopback location, the environmental characteristic of carrier local environment and the fortune of the carrier can also be obtained
Dynamic feature.
In a kind of illustrative specific embodiment, acquired carrier local environment, which obtains environmental characteristic, specifically be can be
The Magnetic Field of the carrier local environment.In practical application, Magnetic Sensor can be installed in the carrier, then build figure process in carrier
In, the Magnetic Field of Magnetic Sensor output can be read.Wherein, the three-dimensional system of coordinate established in carrier itself (or can be claimed
For b system, body coordinate system or carrier coordinate system) in, read Magnetic Field specifically can be in x-axis, y-axis and z-axis
The component in three directions, such as (Mx_b, My_b, Mz_b) etc..
And the motion feature of acquired carrier, it specifically can be carrier pitching angle theta, roll angle γ and course angle Ψ
Deng.Then, accelerometer can be installed in the carrier, in this way, can read to obtain acceleration when obtaining the motion feature of carrier
The ratio force information under b system exported is counted, this is also possible to the component in three x-axis, y-axis and z-axis directions than force information,
Such as (fx_b, fy_b, fz_b) etc..It then, can be according to following formula according to this than force information and acquired Magnetic Field
(1) pitching angle theta, roll angle γ and the course angle Ψ of the carrier are calculated separately out.
Wherein, " g " in formula (1) indicates acceleration of gravity, in a kind of example, g can value be 9.80665m/s2；
" D " is the magnetic declination of carrier position on the earth, can be obtained by searching for certain table.
Further, due to being based on Magnetic Field (Mx_b, the My_ under b system from the read Magnetic Field of Magnetic Sensor
B, Mz_b), and in practical application, when handling the Magnetic Field, it may be necessary to convert the Magnetic Field under the b system
At the Magnetic Field (Mx_n, My_n, Mz_n) under n system (i.e. navigation frame, navigational coordinate system).Then, will be under b system
Magnetic Field be converted into the Magnetic Field under n system during, can first be calculated between two coordinate systems according to formula (2)
Coordinate conversion matrixIt is based on the coordinate conversion matrix againAnd the Magnetic Field (Mx_b, My_b, Mz_b) under b system,
The Magnetic Field (Mx_n, My_n, Mz_n) under n system is calculated according to formula (3).
S202: according to acquired environmental characteristic and motion feature, the winding for being directed to candidate loopback location is calculated
Detect confidence parameter, wherein it is the history winding position that the carrier passes through that winding, which detects the candidate loopback location of confidence parameter characterization,
The confidence level set.
When carrier successively moves to identical position (the history winding that i.e. carrier is presently in position and carrier once passed through
Position is overlapped) when, the motion feature of environmental characteristic and carrier at two same positions of different time points is usually identical, because
This, the motion feature of environmental characteristic and carrier based on carrier local environment can be history winding position to candidate loopback location
A possibility that setting is verified.When specific implementation, in the present embodiment, it is directed to and utilizes candidate's winding position determined by bag of words
It sets, can use acquired environmental characteristic and motion feature to calculate winding detection confidence parameter, and utilize the winding
Detect the confidence level that confidence parameter characterization candidate's loopback location is actual history loopback location.Wherein, confidence level is bigger, shows to wait
A possibility that history loopback location for selecting loopback location once to pass through for carrier, is higher, correspondingly, candidate's loopback location is made
It is also more credible for final winding testing result；Conversely, confidence level is smaller, show that candidate loopback location is history loopback location
A possibility that it is lower, correspondingly, being confirmed as final winding testing result for candidate's loopback location is also more unsuitable.
It, can be according to environmental characteristic and movement as a kind of illustrative embodiments of calculating winding detection confidence parameter
Feature, determine carrier in the quantizating index vector of current location, and obtain quantizating index corresponding to candidate's loopback location to
Amount.It is appreciated that if being really history loopback location (history winding based on the candidate loopback location that bag of words are determined
Position is presently in position with carrier and is overlapped), then carrier current location quantizating index vector and candidate's loopback location institute
Corresponding quantizating index vector should be roughly the same；Conversely, if not being based on the candidate loopback location that bag of words are determined
For history loopback location, then carrier quantizating index corresponding to the quantizating index vector and candidate's loopback location of current location
Vector usually there will be larger difference.Therefore, in the present embodiment, can according to carrier current location quantizating index vector with
And quantizating index vector corresponding to candidate loopback location, the winding detection for being directed to candidate's loopback location can be calculated
Confidence parameter.
As an example, according to carrier in the quantizating index vector of current location and candidate loopback location corresponding to
Quantizating index vector, when calculating winding detection confidence parameter, specifically can be and calculate quantization of the carrier in current location and refer to
The absolute value of the difference of the mould length of each component in quantizating index vector corresponding to vector and the candidate loopback location is marked, so
Afterwards, corresponding winding detection confidence parameter can be calculated further according to the absolute value of the difference of the mould length of each component.
For example, it is assumed that it is Pi that carrier, which is presently in position, then quantizating index vector of the carrier at Pi can be Qi=
[Ψ, Mx_b, My_b, Mz_b, Mx_n, My_n, Mz_n], wherein Ψ is course angle of the carrier at the position Pi.And candidate winding
Quantizating index vector corresponding to the Px of position can be Qx=[Ψ ', Mx_b ', My_b ', Mz_b ', Mx_n ', My_n ', Mz_
N '], wherein Ψ ' is carrier in the corresponding course angle in candidate loopback location place, and Mx_b ', My_b ', Mz_b ' they are respectively carrier
Component of the Magnetic Field in x-axis, y-axis and z-axis direction based on b system when at candidate loopback location, Mx_n ', My_n ', Mz_
N ' be respectively carrier at candidate loopback location when the Magnetic Field based on n system x-axis, y-axis and z-axis direction component.So
Afterwards, the long difference of the mould that each component in quantizating index vector Qi and quantizating index vector Qx can be calculated according to formula (4)
Absolute value, formula (4) is as follows:
Δ Q [k]=| Qi [k] |-| Qx [k] | (4)
Wherein, Δ Q [k] indicates the absolute value of the difference of the mould length of k-th of component in quantizating index vector Qi and Qx, due to
Each quantizating index vector includes 7 components in this example, and therefore, the value of k can be 1 to 7 (in other embodiments, k
The number of value component included in quantizating index vector determine), | Qi [k] | indicate kth in quantizating index vector Qi
The long absolute value of the mould of a component, | Qx [k] | indicate the long absolute value of the mould of k-th of component in quantizating index vector Qx.
Then, corresponding winding can be further calculated out according to formula (5) and formula (6) and detects confidence parameter M.It is public
Formula (5) and formula (6) are as follows:
S203: if winding detection confidence parameter is greater than preset threshold, which is determined as winding detection
As a result.
If being appreciated that, winding detection confidence parameter is larger, is specially greater than preset threshold, shows based on bag of words institute
Determining candidate loopback location is that the confidence level of history loopback location is higher, then can will be determined as candidate's loopback location most
Whole winding testing result；And if winding detection confidence parameter is smaller, is specially not more than the preset threshold, if show should for this
Candidate loopback location is determined as final winding testing result so that the winding testing result there are it is larger a possibility that exist vacation
Therefore positive or false negative in some embodiments, can be refused for candidate's loopback location to be determined as winding detection knot
Fruit, and winding detection can be re-started etc..In this way, detecting confidence parameter using winding, candidate loopback location can be carried out
Certain screening, so that only the higher candidate loopback location of confidence level could be to be determined as winding testing result, and one
The candidate loopback location that false positive or false negative erroneous judgement can be generated with larger possibility a bit can not be confirmed as winding inspection
It surveys as a result, this can effectively improve the accuracy of finally obtained winding testing result, as far as possible reduction False Rate.
In some possible embodiments, the value of preset threshold can be any one numerical value within 0.7 to 1.
Also, when the accuracy requirement for winding testing result is higher, the value of preset threshold can be properly increased, for example, will
Preset threshold can range between 0.9 to 1 carry out value etc..
Further, in practical application, after winding detection position is determined as final winding testing result, carrier
(the SLAM system on carrier) can carry out global pose optimization based on candidate's loopback location.Otherwise, carrier can continue into
Row winding detects and repeats above-mentioned steps, is based on winding detection with the candidate loopback location redefined to bag of words and sets
Letter parameter is screened.
Certainly, if the movement of carrier is not finished, it can repeat the above process and carry out winding detection；And if the movement of carrier
Terminate, then can stop winding detection process, and close SLAM system.
In the present embodiment, the environmental characteristic of available carrier local environment and the motion feature of the carrier, base are obtained
The candidate loopback location of the carrier determined by bag of words, then, according to the environmental characteristic and motion feature, Ke Yijin
One step calculates the winding detection confidence parameter for the candidate loopback location determined based on bag of words, wherein winding inspection
Survey the confidence level for the history loopback location that confidence parameter characterization candidate's loopback location once passed through for the carrier；If winding is examined
It surveys confidence parameter and is greater than preset threshold, then candidate's loopback location is determined as winding testing result.Although as it can be seen that directly by base
It is determined as winding testing result in the candidate loopback location that bag of words obtain carrier, the winding finally determined detection knot can be made
The accuracy of fruit is lower, but according to the environmental characteristic of carrier local environment and the motion feature of carrier, can determine this
Candidate loopback location is the confidence level for the history loopback location that carrier once passed through.It is appreciated that the confidence level is higher, show this
Candidate loopback location be carrier once passed through history loopback location a possibility that it is higher, i.e., carrier be presently in that scene more has can
It can be the scene once passed through, thus it is also higher using the accuracy of candidate's loopback location as the winding testing result, into
And after confidence level higher winding detection position is determined as final winding testing result, it can be improved and finally determined
Winding testing result accuracy.
In addition, the embodiment of the present application also provides a kind of winding detection devices of SLAM system.It is shown refering to Fig. 3, Fig. 3
A kind of winding detection device of SLAM system, the device 300 include: in the embodiment of the present application
Obtain module 301, for obtain the environmental characteristic of carrier local environment and obtain the carrier motion feature,
Based on the candidate loopback location for the carrier that bag of words determine；
Computing module 302, for calculating and being directed to the candidate according to the environmental characteristic and the motion feature
The winding of loopback location detects confidence parameter, and it is the load that the winding, which detects the confidence parameter characterization candidate loopback location,
The confidence level for the history loopback location that body passes through；
Determining module 303, if being greater than preset threshold for winding detection confidence parameter, by the candidate winding position
It sets and is determined as winding testing result.
In some possible embodiments, the motion feature includes the pitch angle, roll angle and boat of the carrier
To angle, the environmental characteristic includes the Magnetic Field of the carrier local environment.
In some possible embodiments, the acquisition module 301, comprising:
First acquisition unit, the ratio force information exported for obtaining the accelerometer being installed on the carrier；
First computing unit, for than force information and the Magnetic Field, calculating bowing for the carrier according to described
The elevation angle, roll angle and course angle.
In some possible embodiments, the computing module 302, comprising:
Determination unit, for determining the carrier described current according to the environmental characteristic and the motion feature
The quantizating index vector of position；
Second acquisition unit, for obtaining quantizating index vector corresponding to the candidate loopback location；
Second computing unit, for according to the carrier in the current location quantizating index vector and the candidate
The winding detection confidence ginseng for being directed to the candidate loopback location is calculated in quantizating index vector corresponding to loopback location
Number.
In some possible embodiments, second computing unit, comprising:
First computation subunit, for calculating quantizating index vector and the time of the carrier in the current location
Select the absolute value of the difference of the mould length of each component in quantizating index vector corresponding to loopback location；
Second computation subunit is calculated and is directed to for the absolute value of the difference according to the mould length of each component
The winding of candidate's loopback location detects confidence parameter.
It is worth noting that, the winding detection device of SLAM system described in the present embodiment, it is real to correspond to the above method
The winding detection method of SLAM system described in example is applied, therefore, specific embodiment can be refering to the phase of embodiment of the method
Place description is closed, this will not be repeated here.
In the present embodiment, although the candidate loopback location for obtaining carrier based on bag of words is directly determined as winding detection
As a result, the accuracy of the winding testing result finally determined can be made lower, but according to the environmental characteristic of carrier local environment
And the motion feature of carrier, it can determine that candidate's loopback location is the confidence for the history loopback location that carrier once passed through
Degree.It is appreciated that the confidence level is higher, show that candidate's loopback location is the possibility for the history loopback location that carrier once passed through
Property is higher, i.e. carrier is presently in scene and is more likely to be the scene once passed through, thus using candidate's loopback location as
The accuracy of winding testing result is also higher, in turn, confidence level higher winding detection position is determined as final winding
After testing result, the accuracy for the winding testing result finally determined can be improved.
" first acquisition unit " mentioned in the embodiment of the present application, " the first computing unit ", " the first computation subunit " etc.
" first " in title is used only to do name mark, does not represent first sequentially.The rule is equally applicable to " second "
Deng.
As seen through the above description of the embodiments, those skilled in the art can be understood that above-mentioned implementation
All or part of the steps in example method can add the mode of general hardware platform to realize by software.Based on this understanding,
The technical solution of the application can be embodied in the form of software products, which can store is situated between in storage
In matter, such as read-only memory (English: read-only memory, ROM)/RAM, magnetic disk, CD etc., including some instructions to
So that a computer equipment (can be the network communication equipments such as personal computer, server, or router) executes
Method described in certain parts of each embodiment of the application or embodiment.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device reality
For applying example, since it is substantially similar to the method embodiment, so describing fairly simple, related place is referring to embodiment of the method
Part explanation.The apparatus embodiments described above are merely exemplary, wherein mould as illustrated by the separation member
Block may or may not be physically separated, and the component shown as module may or may not be physics
Module, it can it is in one place, or may be distributed over multiple network units.It can select according to the actual needs
Some or all of the modules therein achieves the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creation
Property labour in the case where, it can understand and implement.
The above is only the illustrative embodiment of the application, is not intended to limit the protection scope of the application.
Claims (10)
1. a kind of winding detection method of SLAM system, which is characterized in that the described method includes:
The institute for obtaining the environmental characteristic of carrier local environment and obtaining the motion feature of the carrier, being determined based on bag of words
State the candidate loopback location of carrier；
According to the environmental characteristic and the motion feature, calculates and be directed to the winding detection of the candidate loopback location and set
Believe parameter, it is the history loopback location that the carrier passes through that the winding, which detects the confidence parameter characterization candidate loopback location,
Confidence level；
If the winding detection confidence parameter is greater than preset threshold, the candidate loopback location is determined as winding detection knot
Fruit.
2. the method according to claim 1, wherein the motion feature includes the pitch angle of the carrier, cross
Roll angle and course angle, the environmental characteristic include the Magnetic Field of the carrier local environment.
3. according to the method described in claim 2, it is characterized in that, the motion feature for obtaining carrier, comprising:
Obtain the ratio force information that the accelerometer being installed on the carrier is exported；
According to described than force information and the Magnetic Field, the pitch angle, roll angle and course angle of the carrier are calculated.
4. the method according to claim 1, wherein described special according to the environmental characteristic and the movement
Sign calculates the winding detection confidence parameter for being directed to the candidate loopback location, comprising:
According to the environmental characteristic and the motion feature, determine the carrier the current location quantizating index to
Amount；
Obtain quantizating index vector corresponding to the candidate loopback location；
According to carrier quantization corresponding to the quantizating index vector and the candidate loopback location of the current location
The winding detection confidence parameter for being directed to the candidate loopback location is calculated in indicator vector.
5. according to the method described in claim 4, it is characterized in that, the quantization according to the carrier in the current location
Quantizating index vector corresponding to indicator vector and the candidate loopback location, is calculated and is directed to the candidate winding position
The winding detection confidence parameter set, comprising:
Calculate carrier quantization corresponding to the quantizating index vector and the candidate loopback location of the current location
The absolute value of the difference of the mould length of each component in indicator vector；
According to the absolute value of the difference of the mould length of each component, the winding inspection for being directed to the candidate loopback location is calculated
Survey confidence parameter.
6. a kind of winding detection device of SLAM system, which is characterized in that described device includes:
Module is obtained, for obtaining the environmental characteristic of carrier local environment and obtaining the motion feature, word-based of the carrier
The candidate loopback location for the carrier that bag model determines；
Computing module is directed to the candidate winding position for calculating according to the environmental characteristic and the motion feature
The winding detection confidence parameter set, it is carrier process that the winding, which detects the confidence parameter characterization candidate loopback location,
History loopback location confidence level；
Determining module determines the candidate loopback location if being greater than preset threshold for winding detection confidence parameter
For winding testing result.
7. device according to claim 6, which is characterized in that the motion feature includes the pitch angle of the carrier, cross
Roll angle and course angle, the environmental characteristic include the Magnetic Field of the carrier local environment.
8. device according to claim 7, which is characterized in that the acquisition module, comprising:
First acquisition unit, the ratio force information exported for obtaining the accelerometer being installed on the carrier；
First computing unit, for than force information and the Magnetic Field, calculated according to described the carrier pitch angle,
Roll angle and course angle.
9. device according to claim 6, which is characterized in that the computing module, comprising:
Determination unit, for determining the carrier in the current location according to the environmental characteristic and the motion feature
Quantizating index vector；
Second acquisition unit, for obtaining quantizating index vector corresponding to the candidate loopback location；
Second computing unit, for the quantizating index vector and the candidate winding according to the carrier in the current location
The winding detection confidence parameter for being directed to the candidate loopback location is calculated in quantizating index vector corresponding to position.
10. device according to claim 9, which is characterized in that second computing unit, comprising:
First computation subunit is returned for calculating the carrier in the quantizating index vector of the current location and the candidate
The absolute value of the difference of the mould length of each component in quantizating index vector corresponding to ring position；
Second computation subunit, for the absolute value of the difference according to the mould length of each component, be calculated be directed to it is described
The winding of candidate loopback location detects confidence parameter.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910804329.6A CN110514199B (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
| CN202110962668.4A CN113670300A (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910804329.6A CN110514199B (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110962668.4A Division CN113670300A (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110514199A true CN110514199A (en) | 2019-11-29 |
| CN110514199B CN110514199B (en) | 2021-10-22 |
Family
ID=68627654
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110962668.4A Pending CN113670300A (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
| CN201910804329.6A Active CN110514199B (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110962668.4A Pending CN113670300A (en) | 2019-08-28 | 2019-08-28 | Loop detection method and device of SLAM system |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN113670300A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024007807A1 (en) * | 2022-07-06 | 2024-01-11 | 杭州萤石软件有限公司 | Error correction method and apparatus, and mobile device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780608A (en) * | 2016-11-23 | 2017-05-31 | 北京地平线机器人技术研发有限公司 | Posture information method of estimation, device and movable equipment |
| CN108303099A (en) * | 2018-06-14 | 2018-07-20 | 江苏中科院智能科学技术应用研究院 | Autonomous navigation method in unmanned plane room based on 3D vision SLAM |
| CN108665540A (en) * | 2018-03-16 | 2018-10-16 | 浙江工业大学 | Robot localization based on binocular vision feature and IMU information and map structuring system |
| CN109409418A (en) * | 2018-09-29 | 2019-03-01 | 中山大学 | A kind of winding detection method based on bag of words |
| CN109682385A (en) * | 2018-11-05 | 2019-04-26 | 天津大学 | A method of instant positioning and map structuring based on ORB feature |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108805149A (en) * | 2017-05-05 | 2018-11-13 | 中兴通讯股份有限公司 | A kind of winding detection method and device of visual synchronization positioning and map structuring |
| CN109995799B (en) * | 2017-12-29 | 2020-12-29 | Oppo广东移动通信有限公司 | Information pushing method and device, terminal and storage medium |
| CN108253958B (en) * | 2018-01-18 | 2020-08-11 | 亿嘉和科技股份有限公司 | Robot real-time positioning method in sparse environment |
| CN110044354B (en) * | 2019-03-28 | 2022-05-20 | 东南大学 | Binocular vision indoor positioning and mapping method and device |
-
2019
- 2019-08-28 CN CN202110962668.4A patent/CN113670300A/en active Pending
- 2019-08-28 CN CN201910804329.6A patent/CN110514199B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780608A (en) * | 2016-11-23 | 2017-05-31 | 北京地平线机器人技术研发有限公司 | Posture information method of estimation, device and movable equipment |
| CN108665540A (en) * | 2018-03-16 | 2018-10-16 | 浙江工业大学 | Robot localization based on binocular vision feature and IMU information and map structuring system |
| CN108303099A (en) * | 2018-06-14 | 2018-07-20 | 江苏中科院智能科学技术应用研究院 | Autonomous navigation method in unmanned plane room based on 3D vision SLAM |
| CN109409418A (en) * | 2018-09-29 | 2019-03-01 | 中山大学 | A kind of winding detection method based on bag of words |
| CN109682385A (en) * | 2018-11-05 | 2019-04-26 | 天津大学 | A method of instant positioning and map structuring based on ORB feature |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024007807A1 (en) * | 2022-07-06 | 2024-01-11 | 杭州萤石软件有限公司 | Error correction method and apparatus, and mobile device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110514199B (en) | 2021-10-22 |
| CN113670300A (en) | 2021-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10341982B2 (en) | Technique and system of positioning a mobile terminal indoors | |
| CN109270545B (en) | Positioning true value verification method, device, equipment and storage medium | |
| CN109579847A (en) | Extraction method of key frame, device and smart machine in synchronous superposition | |
| CN109540148A (en) | Localization method and system based on SLAM map | |
| US10970541B2 (en) | Geo-registering an aerial image by an object detection model using machine learning | |
| EP3789797B1 (en) | Method and apparatus for evaluating data, device, and computer-readable storage medium | |
| Miao et al. | UniVIO: Unified direct and feature-based underwater stereo visual-inertial odometry | |
| Ishihara et al. | Beacon-guided structure from motion for smartphone-based navigation | |
| CN115164936A (en) | Global pose correction method and device for point cloud splicing in high-precision map manufacturing | |
| Varelas et al. | An AR indoor positioning system based on anchors | |
| CN110514199A (en) | A kind of winding detection method and device of SLAM system | |
| Forno et al. | Techniques for improving localization applications running on low-cost IoT devices | |
| CN113218380B (en) | Electronic compass correction method and device, electronic equipment and storage medium | |
| Praschl et al. | Enabling outdoor MR capabilities for head mounted displays: a case study | |
| CN113375665B (en) | Unmanned aerial vehicle pose estimation method based on multi-sensor elastic coupling | |
| CN113917506A (en) | Ambiguity fixing method and device, electronic equipment and automatic driving equipment | |
| Wang et al. | Pedestrian positioning in urban city with the aid of Google maps street view | |
| Karan et al. | Indoor and outdoor navigation application development with augmented reality technology | |
| US20210131808A1 (en) | Method for reconstructing the movement of an individual and the signal map of a location | |
| Huang et al. | Indoor navigation with iPhone/iPad: Floor plan based monocular vision navigation | |
| Inoue et al. | Markovian jump linear systems-based filtering for visual and GPS aided inertial navigation system | |
| Liu et al. | The design of an embedded multi-sensor data fusion system for unmanned surface vehicle navigation based on real time operating system | |
| Björklund | Tracking Multiple Objects Using Computer Vision and Inertial Measurement Unit | |
| CN117611762B (en) | Multi-level map construction method, system and electronic equipment | |
| CN110398765A (en) | Localization method and device, unmanned 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |