CN110296707A - Indoor service robot based on VLC technology - Google Patents
Indoor service robot based on VLC technology Download PDFInfo
- Publication number
- CN110296707A CN110296707A CN201910702032.9A CN201910702032A CN110296707A CN 110296707 A CN110296707 A CN 110296707A CN 201910702032 A CN201910702032 A CN 201910702032A CN 110296707 A CN110296707 A CN 110296707A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- robot
- led
- service robot
- 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.)
- Withdrawn
Links
- 238000005516 engineering process Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 64
- 238000004891 communication Methods 0.000 claims abstract description 44
- 238000003384 imaging method Methods 0.000 claims abstract description 24
- 230000000694 effects Effects 0.000 claims abstract description 21
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 230000003287 optical effect Effects 0.000 claims abstract description 15
- 230000008569 process Effects 0.000 claims description 16
- 230000033001 locomotion Effects 0.000 claims description 14
- 230000000007 visual effect Effects 0.000 claims description 13
- 230000005540 biological transmission Effects 0.000 claims description 12
- 230000004807 localization Effects 0.000 claims description 10
- 239000002245 particle Substances 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 8
- 230000007246 mechanism Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 5
- 238000005286 illumination Methods 0.000 claims description 4
- 238000003709 image segmentation Methods 0.000 claims description 4
- 238000007630 basic procedure Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000011056 performance test Methods 0.000 claims description 3
- 238000013112 stability test Methods 0.000 claims description 3
- 239000012141 concentrate Substances 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 18
- 230000007613 environmental effect Effects 0.000 description 7
- 230000004438 eyesight Effects 0.000 description 7
- 241001465382 Physalis alkekengi Species 0.000 description 6
- 230000008901 benefit Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008447 perception Effects 0.000 description 6
- 238000001514 detection method Methods 0.000 description 5
- 230000009172 bursting Effects 0.000 description 4
- 230000008033 biological extinction Effects 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005021 gait Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000001771 impaired effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 229910044991 metal oxide Inorganic materials 0.000 description 1
- 150000004706 metal oxides Chemical class 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 231100000572 poisoning Toxicity 0.000 description 1
- 230000000607 poisoning effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 235000015170 shellfish Nutrition 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000016776 visual perception Effects 0.000 description 1
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
- G01C21/20—Instruments for performing navigational calculations
- G01C21/206—Instruments for performing navigational calculations specially adapted for indoor navigation
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Optical Communication System (AREA)
Abstract
A kind of indoor service robot based on VLC, indoor service robot based on VLC carries out indoor positioning using visible light communication technology, following three kinds of sensors are installed in robot, that is CMOS camera, kinect sensor and inertial sensor, write the circuit of driver respectively on these three sensors, and the positioning system with indoor service robot is positioned respectively, it passes through roller shutter effect and just word length coding techniques, pass through optical flow method, pixel coordinate and image coordinate are converted, it is final to obtain Indoor Robot product, compared to other location technologies based on wireless communication, the location algorithm that the application proposes is established on existing lighting system, without adding any specific sending module, and realize that high-precision imaging positions using the camera that service robot carries, increasing While adding positioning accuracy, the input cost of system is reduced, meets the requirement of economizing economy.
Description
Technical field
The application belongs to robot technology, more particularly, to a kind of indoor service robot.
Background technique
In the correlative study and exploitation of service robot, it is that robot is smoothly complete that independent navigation, which is a core technology,
Key points and difficulties in one of basic link at various services and mobile robot research field, although positioning side at present
Method is many kinds of, and various algorithms are also to emerge one after another, but to realize the reliable of robot and be accurately positioned, and still there are many ask
Topic needs to solve, such as the limitation of environmental characteristic, the deviation accumulation of related algorithm, the barrier based on complicated landform pass through
Deng, and indoor environment has the characteristics that limitations, indoor light illumination contrast are larger, these features all affect positioning side
The selection of method.
Domestic and international indoor moving service robot location and navigation technology present Research:
Indoor moving service robot is mainly used under a kind of working environment of structuring, such as family, hospital, navigation system
It is indispensable pith in service robot, its major issue to be solved is:
1, accurate detection of the robot in the position in space, direction, environmental information；
2, the analysis of obtained information and the foundation of environmental model；
3, the trajectory path planning for making robot security mobile, when running under unknown and uncertain environment, mobile robot must
Must by sensor collection data, and by certain algorithm sensor collection to data analyzed, merged for building
The model of vertical external environment provides correct foundation for navigation decision.
Existing indoor service robot localization airmanship mainly include the following types:
(1) based on the location technology of reckoning；
Reckoning does not need external sensor information to realize the estimation to vehicle location and direction, and is capable of providing very high
Short-term positioning accuracy.The key of reckoning technology be mobile robot unit interval be can be measured pass by away from
From, and the during this period of time variation in mobile robot course, this method have the advantages that it is self-contained, i.e., without external ginseng
It examines, however, there is drift at any time, after integral, any small constant error can all increase without limitation, therefore, for prolonged
Accurate positioning is not applicable.
(2) based on the localization method of signal lamp:
Signal lamp positioning system is the navigator fix means that ship and aircraft generally use, based on the positioning system of signal lamp according to
Rely one group of installation known signal lamp in the environment, sensor is installed in mobile robot, signal lamp is observed, is used for
There are many kinds of the sensors of environmental observation, can be the signal of active, such as active vision, ultrasonic wave, laser radar, can also be with
It is passively signal, such as GPS, although this localization method provides very high sample rate and high stability, peace
Dress and maintenance beacon cost are very high.
(3) based on the localization method of map:
In the location technology based on map, map structuring is one of them important content, currently mainly has topological structure to retouch
State map and two kinds of geometry map:
The topological map level of abstraction is high, there is following advantage:
1. being conducive to further path and mission planning；
2. storage and search space are all smaller, computational efficiency is high；
3. many existing mature, efficient search and reasoning algorithm can be used, disadvantage: topological map is a figure
(Graph), by node and Bian Zucheng, only consider the connectivity between node, without considering path planning, it relaxes map to essence
The needs of true position are bad at the map that expression has labyrinth, and how to map is split to form node and side, for another example
What carries out navigation and path planning using topological map, is still the problem of requiring study.
Geometry map can be grid description, be also possible to line segment either polygon description, advantage be establish
It is easy, remains the various information of entire environment as far as possible, the identification to environmental characteristic is also no longer dependent in position fixing process, but
It is that search space is very big in position fixing process, if being just difficult to meet requirement of real-time without preferably simplifying algorithm, grid
Figure constructs an accurate environmental map, be registrated by the information to observation with map can calculating robot position
It sets.
(4) based on the localization method of road sign:
Road sign is the different characteristics that robot can recognize from its sensing input, and road sign can be geometry, as line segment,
Circle, rectangle, may also comprise additional information, ordinary circumstance, road sign has fixed and known position, and road sign conscientiously will be selected carefully
Select, in favor of identification, obtain problem to simplify road sign, always assume that the position of current robot and orientation it is approximate it is known that this
Sample can make robot find road sign in a limited region, its advantage is that being conveniently easily achieved, positioning accuracy is relatively
It is high.However target rational deployment of satisfying the need but being needed, and installation is big, maintenance cost is high.
(5) localization method of view-based access control model:
The positioning and directing of view-based access control model realizes perception and understanding to environment using computer vision technique, analyzes road
Structure, identify the current roadway area of positioning, and then make roading in real time according to mission requirements, monitor and drive and drive
It sails device and executes this planning, make it, have benefited from digital image processing techniques relative maturity, the positioning side of view-based access control model
The environmental information that method is got is abundant.But visual perception is affected by sight and light, is wanted to processor processing speed
Ask high.
Summary of the invention
The technical issues that need to address:
1, the problem of low influence of code efficiency of imaging positioning system positions real-time:
Communication efficiency directly affects the real-time of VLC positioning system, although and the real-time of positioning system influences static immobilization
It is less, but be very significantly that high code efficiency can greatly improve communication efficiency, be based on to dynamically track positioning effects
VLC is imaged in positioning system, and the coding methods such as existing 4-PPM, template matching method and frequency coding are low etc. there are code efficiency
Problem.
2, effect of bursting forth in visual communication and influence communication decoding the problem of:
For image obtained in visual communication, can be excessively high due to brightness in LED pixel point intermediate region and generate bloom and overflow
Out, the phenomenon that leading to that electric charge transfer occurs between two neighboring pixel, causing the contrast reduction at bright and dark edge, from
And lead to the problem of cannot normally interpreting light and shade striped coding.
3, the real time problems of the service robot tracking and positioning of view-based access control model:
It realizes the real-time positioning of Information Mobile Service robot, just captured in real-time and each image must be handled, find out lamps and lanterns image,
Encoded information is extracted, calculates the real time position of Information Mobile Service robot, CMOS camera opens image pickup mode can captured in real-time
Image, when the camera random device people on service robot is mobile, position of the image of lamps and lanterns in sequence of video images
Also in relative motion, if carrying out lamps and lanterns searching for every frame image as single image, such each image will be carried out once
Complete search, calculation amount is huge, and real-time is poor, even and if being also difficult to accurately track every width in real time with the camera of high speed
The position of LED in image.
Technical solution:
1, a kind of indoor service robot based on VLC, it is characterized in that: the indoor service robot based on VLC uses visible light
The communication technology carries out indoor positioning, installs following three kinds of sensors in robot, i.e., CMOS camera, kinect sensor with
And inertial sensor, write the circuit of driver respectively on these three sensors, and respectively with indoor service robot
Positioning system positioned
CMOS camera working principle is: CMOS camera installs start-up circuit, and obtains compressed data using opencv, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and synchronizes and is decoded, later, utilizes video
Driving, is then decoded image, is detected by LED light, obtain the ID number of LED light, so that the position for obtaining robot is sat
Mark, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Kinect sensor working principle is: Kinect sensor installs start-up circuit, and is obtained using cartographer
Compressed data, then compressed data is transmitted to by network or WIFI the transport layer of positioning system, and synchronize and be decoded, it
Afterwards, using video drive, then image is decoded, by feature extraction, instant map is created, to obtain robot
Position coordinates, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Inertial sensor working principle is: inertial sensor installs start-up circuit, and obtains metrical information using Kalman, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and signal processing is carried out to metrical information, it
Afterwards, posture information is updated, then by reckoning, the estimation of the position coordinates of robot is carried out based on particle filter algorithm.
Indoor service robot according to claim 1 based on VLC, it is characterized in that: being obtained by following steps
Robot product:
Step 1: carrying out theoretical calculation first with VLC and visual identity and carry out digital modeling using computer modeling software；
Step 2: be respectively to obtain the numerical chracter of visible light communication using following three kinds of technologies:
2.1: utilizing the imaging mechanism of CMOS camera；
2.2: utilizing the mB/nB variable length code technology of VLC；
2.3: utilizing the imaging Location Mechanism of single camera；
Step 3: by optical flow method and bayesian criterion, realizing the tracking to LED lamp track；
Step 4: the big positioning system of fusion following two, and VLC movable object tracking positioning system is developed, two big positioning system difference
It is:
4.1:cartographer open source SLAM positioning system；
4.2: inertial positioning system；
Step 5: obtaining high-accuracy position system in the service robot room based on visible light communication；
Step 6: positioning performance test and stability test；
Step 7: final to obtain service robot product.
Indoor service robot according to claim 1 based on VLC, it is characterized in that: the acquisition LED lamp
The ID location information transmitted, basic procedure is:
The image of intake is obtained a frame image of caching by step 1, camera；
Step 2 pre-processes the image of caching；
Step 3 carries out downlight image segmentation to image；
Step 4, the extraction that downlight feature is carried out to image；
Step 5 does not include downlight, then gives up image, includes downlight, into next step；
Step 6, the selection into striate region；
Step 7 pre-processes striate region；
Whether step 8, analysis striped include that two color heads are given up if do not included, if comprising into next step；
Step 9, carry out color verification, color verification mainly color difference is verified, in RGB equipment, due to display screen with
The difference of camera lens manufacturing process and external visible light interference, cause the identical color of binary data, by showing, shooting with video-corder
After technique, color difference will result in, still, color difference is acceptable to a certain extent；
Step 10, by analysis, sees whether the data of transmission are correct, if incorrect, give up, if correctly, executing solution
Code.
Indoor service robot according to claim 1 based on VLC, it is characterized in that: the positioning system
Positioning the following steps are included:
Step 1, camera are located on service robot, to camera positioning namely to robot localization,
The CMOS camera being mounted on service robot opens image pickup mode, in the transmitting of variable length code technology and receiving
Mentioned detects the position of LED lamp and determines to take in initial image frame based on visible light communication imaging location technology
It is engaged in the initial position of robot, however when CMOS camera is moved relative to LED lamp, LED lamp on the image can be due to
Movement generates blurring effect, and due to the interference of environment light, selects so as to cause LED pixel and be blended in one with environment light image vegetarian refreshments
It rises, therefore enters step 2；
Step 2, visual optic communication are all based on institute's captured image image pixel intensities to distinguish LED signal point and background dot picture
Element calculates the conditional probability that a certain pixel is LED signal point by obtaining the image pixel intensities of image, it is assumed that n-th frame image
Ith pixel point intensity be f1 (n), wherein f1 (n)With Gaussian Profile；
Step 3 goes out LED pixel moving direction light stream figure in image, and then the service of obtaining by the continuous frame picture construction of two width
Robot moving direction is since the time interval of adjacent two field pictures is very short, and the image pixel intensities of LED are several on two inter frame images
It is consistent, passes through the locality of the position of present frame LED pixel point LED pixel point corresponding with previous frame and the change of amplitude
Change, it can determine the moving condition of pixel, wherein effective movement pixel passes through setting position direction and amplitude function
Threshold value, when be lower than threshold value, it is believed that be stable；Higher than threshold value, just effectively to move pixel, calculate each in image
After the effective mobile vector of LED pixel, the motion state of service robot can be obtained, this step is optical flow method；
Step 4, using the feature of image produced by step 2 and step 3 as class conditional probability density, further according to Bayesian formula
The posterior probability that each pixel in image belongs to background dot or LED signal point is calculated, by the pixel on every frame image
It is divided into two classes: W1Class, that is, LED lamp and W2Class, that is, background pixel is classified with probability function P ∈ [0,1], will be by above-mentioned light
θ, r and f of stream method and the collected pixel of image pixel intensities1 (n)All it is recorded as in f1 (n), as class conditional probability density
Function, and in the (n-1)th frame image, the number of LED pixel point is divided by the number probability obtained of total pixel as first
Probability is tested, for n-th frame, W1Class and W2The posterior probability of class is by Bayesian formula by class conditional probability density and the (n-1)th frame figure
The prior probability of picture calculates, and judges that each pixel belongs to background dot still in n-th frame image by Bayes rule
LED signal point.
Indoor service robot according to claim 2 based on VLC, it is characterized in that: the mB/nB mutilation is long
The technical solution of coding is: former word length is the symbol of m, if comprising 0 and 1 several difference be greater than 2, be n from word length
A symbol is selected in code-element set so that comprising 0 and 1 several difference be not more than 2, substitute original symbol and transmitted；
Former word length be m symbol, if comprising 0 and 1 several difference be not more than 2, still transmitted with m symbols, change
The purpose of word length coding techniques is to obtain higher illumination stability simultaneously, improves code efficiency,
It comprises the concrete steps that:
Step 1, initial data pass through the long color coding of mB/nB mutilation；
Step 2, driving LED light, to obtain specified color information value；
Step 3, camera obtain its color information value；
Step 4 passes through decoding, obtains new color data.
The utility model has the advantages that the location algorithm that the application proposes is established compared to other location technologies based on wireless communication
On existing lighting system, without adding any specific sending module, and it is real using the camera that service robot carries
Existing high-precision imaging positioning, while increasing positioning accuracy, reduces the input cost of system, meets wanting for economizing economy
It asks.
Detailed description of the invention
Attached drawing 1 is 2D grating map and 2D topological map schematic diagram in Orientation on map；
Attached drawing 2 is technical solution implementation flow chart；
Attached drawing 3 is the principle analysis figure of roller shutter effect；
Attached drawing 4 is the effect diagram of effect of bursting forth；
Attached drawing 5 is the transmitting of variable length code technology and receiving end block diagram；
Attached drawing 6 is color coding schematic diagram；
Attached drawing 7 is image recognition and decoding process schematic diagram；
Attached drawing 8 is that indoor visible light monocular vision positions schematic diagram；
Attached drawing 9 is the geometrical light-path figure of visible light monocular vision positioning；
Attached drawing 10 is pixel coordinate and image coordinate Transformation Graphs；
Attached drawing 11 is the building schematic diagram of the mobile vector based on optical flow method；
Attached drawing 12 is the visible light communication indoor locating system schematic diagram based on ROS service robot；
Attached drawing 13 is the visible light communication positioning system schematic diagram of view-based access control model perception；
Attached drawing 14 is the service robot visible light communication positioning system schematic diagram in true environment；
Attached drawing 15 is service robot positioning system flow chart.
Specific embodiment
Applicant is explained its specific embodiment referring to attached drawing:
Attached drawing 1 is 2D grating map and 2D topological map schematic diagram in Orientation on map, due to 2D grating map and 2D topology ground
Figure is positioned as the relevant technologies in background technique, its mechanism is no longer described in detail in applicant.
Attached drawing 2 illustrates technical solution basic flow chart, is respectively:
1: carrying out theoretical calculation first with VLC and visual identity and carry out digital modeling using computer modeling software；
2: be respectively to obtain the numerical chracter of visible light communication using following three kinds of technologies:
2.1: utilizing the imaging mechanism of CMOS camera；
2.2: utilizing VLC variable length code technology；
2.3: utilizing the imaging Location Mechanism of single camera；
3: by optical flow method and bayesian criterion, realizing the tracking to LED lamp track；
4: the big positioning system of fusion following two, and VLC movable object tracking positioning system is developed, two big positioning systems are respectively:
4.1:cartographer open source SLAM positioning system；
4.2: inertial positioning system；
5: obtaining high-accuracy position system in the service robot room based on visible light communication；
6: positioning performance test and stability test；
7: final to obtain service robot product.
Attached drawing 3 is the principle analysis figure of roller shutter effect:
The frame rate of general existing commercial camera is about 30 ~ 60Hz, and the frame rate refers to one second for 30 ~ 60Hz cuts
The picture number changed is 30 ~ 60 width, if only capturing LED light on and off information by a picture, limited frame rate can the big day of one's doom
The traffic rate of VLC is made, so the rate of VLC communication must be improved by the roller shutter effect of CMOS camera, to protect
Having demonstrate,proved the low frame rate of camera also can receive the optic communication signal of high speed, CMOS, that is, complementary metal oxide semiconductor,
The timing of image is captured as shown in figure 3, general existing commercial its core component of camera is CCD, for ccd image sensor
For, entire image can expose simultaneously, and all pixel datas can also be read simultaneously, and for cmos image sensor, lead to
The mode of progressive scan is crossed to handle the pixel of every a line, integrates the time of a frame image from the first row pixel of imaging sensor
It resets exposure to start, one-row pixels reading to the end terminates, and in this stage, the quick light on and off of LED can generate bright on the image
The data for detecting multidigit on a frame picture may be implemented by roller shutter effect, and then greatly improve communication for dark striped
Rate.
Attached drawing 4 is the effect schematic diagram that bursts forth:
It is decoded simultaneously as in image where LED center in the light and shade striped of the image captured using roller shutter effect
Region point light intensity is higher, is saturated spilling so as to cause the electronics of position pixel, can seriously affect light and shade in every frame image
The contrast of striped is shown in the line segment in Fig. 4 where B, due to central point mistake if directly choosing the corresponding column pixel of LED center point
It is bright and some dark fringes is caused also to become bright fringes, so as to cause that cannot decode or occur serious distortion, that is, effect of bursting forth meeting
Cause to decode or occur serious distortion, so, the application is more using second order by the pixel to every a line in image
Item formula fitting, every piece image obtained first carry out the grayvalue transition of 0 ~ 255 rank, then set in every one-row pixels unit
The row ith pixel is the gray value of respective pixel, resets as second order polynomial fit curve.It is obtained by setting threshold value
Interception area between second order polynomial fit curve and gray value curve, to obtain one group of gray value, then these gray values
Arranged by ascending order, select 20% when value, when detection is dark fringe, selected gray value will be close to 0；Work as inspection
When being bright fringes, selected gray value can be much larger than 0 for survey, and then obtain the gray value of a column matrix to be decoded,
Extinction ratio for obtaining the gray value of a column matrix its light and shade striped generally will not be too high, to greatly limit transmission
Distance and anti-bias light interference performance, therefore, by histogram equalization by the pixel transform of image be evenly distributed on
256 rank gray levels enhance the interfringe extinction ratio of light and shade, recycle Sobel filter to further enhance extinction ratio, most
Afterwards, it is fitted using three rank multinomials a suitable gray threshold detection gray value is arranged to go back the position ID of original encoding and believes
Breath.
Attached drawing 5 is the transmitting of variable length code technology and record principle schematic diagram:
It being communicated relative to high speed VLC, VLC indoor locating system is an application of the VLC in low speed field, so general use
The relatively simple OOK modulator approach of circuit structure, most of communication code methods based on visual light imaging positioning system are also
It developed on the basis of OOK modulator approach, coding method, template matching method, frequency coding method such as based on data flow,
However there is communication code low efficiency in it.So the application proposes a kind of long color coding techniques of mB/nB mutilation, mB/
The technical solution of nB variable length code is: former word length be m symbol, if comprising 0 and 1 several difference be greater than 2, from word
A symbol is selected in the code-element set of a length of n so that comprising 0 and 1 several difference be not more than 2, substitute original symbol into
Row transmission；Former word length be m symbol, if comprising 0 and 1 several difference be not more than 2, still carried out with m symbols
Transmission, the purpose of variable length code technology are to obtain higher illumination stability simultaneously, improve code efficiency,
It comprises the concrete steps that:
1, initial data passes through the long color coding of mB/nB mutilation；
2, LED light is driven, to obtain specified color information value；
3, camera obtains its color information value；
4, by decoding, new color data is obtained.
The transmitting of variable length code technology implements simple, at low cost advantage with hardware with it is received.
Attached drawing 6 is color coding schematic diagram；
Further, byte-header or specificator (including starting character, separator and full stop) usually can be added in communication, improved
Anti-interference ability, while as receiving end start bit.Common method is to add one or more special byte in data word
Before section, however the redundancy for sending data is but increased, reduces the efficiency of transmission.So the application proposes to use face
Color coding techniques, by using field sequence color display technology using colour bits as preposition head, for showing the beginning of byte.Scheming
Shown in 6, A0A1A2 are preposition heads, and for determining the initial position of data, A3 to A16 is data bit, indicate transmission rate position
A length of 14.The long color coding techniques of mutilation can complete complete data with a frame image and transmit, and include byte-header and data
Position long enough, it should be noted that in attached drawing 6, the part of black interval indicates that bright fringe, black region indicate dark fringe, this
In be related to the stability problem of a data transfer problem and data, in the picture, if indicating 1 with bright fringe, dark fringe
It indicates 0, is exactly 01 character string after data conversion, image transmitting has more stable anti-interference ability, for example, 01 data pass
When defeated, it is only necessary to partial data region is damaged, and data transmission fault can be caused, and when image transmitting, topography
Damage can correctly transmit data as long as impaired area is no more than certain degree.
Attached drawing 7 is image recognition and its decoding process schematic illustration:
The extraction of ID location information
The process of visual imaging positioning includes two processes of extraction and visual token of ID location information.CMOS camera is in institute
The number (namely ID location information) for extracting lamps and lanterns in the image information that LED lamp issues is captured, needs to carry out image recognition.
Image recognition is that image is classified to treated.The light and shade as caused by roller shutter effect is extracted on the basis of image segmentation
Striped obtains the ID location information that LED lamp is transmitted, basic procedure is according to the decoding of light and shade streak feature:
The image of intake is obtained a frame image of caching by step 1, camera；
Step 2 pre-processes the image of caching；
Step 3 carries out downlight image segmentation to image；
Step 4, the extraction that downlight feature is carried out to image；
Step 5 does not include downlight, then gives up image, includes downlight, into next step；
Step 6, the selection into striate region；
Step 7 pre-processes striate region；
Whether step 8, analysis striped include that two color heads are given up if do not included, if comprising into next step；
Step 9, carry out color verification, color verification mainly color difference is verified, in RGB equipment, due to display screen with
The difference of camera lens manufacturing process and external visible light interference, cause the identical color of binary data, by showing, shooting with video-corder
After technique, color difference will result in, still, color difference is acceptable to a certain extent；
Step 10, by analysis, sees whether the data of transmission are correct, if incorrect, give up, if correctly, executing solution
Code.
Attached drawing 8 is indoor visible light monocular vision positioning schematic diagram:
In Fig. 8, three LED lamps, which are ceiling mounted, transmits its three-dimensional coordinate information, the CMOS camera shooting on service robot
Head is received and is demodulated coordinate information by roller shutter effect, for receiving the imaging sensor of visible light signal by two-dimensional light sensitive element
Array composition.Therefore, each pixel can be used as an individual sensor, can be detected simultaneously by single image sensor and
Demodulate multiple LED signals.Imaging sensor receives the optical signal of three LED, demodulates to coordinate information, lens it is vertical
The center of axis correspondence image sensor.P point is the center of lens.
The distance of P point and A, B, C are d1, d2, d3.Distance d be by the focal length of lens and LED on the image sensor
What the geometrical relationship between position was calculated.
Fig. 9 is the geometrical light-path schematic diagram of visible light monocular vision positioning:
In the plane of imaging sensor, the distance of LED image center and image sensor center is by obtaining image sensing
Distance in device plane between available two pixels of two pixels, and the preimage LED lamp of corresponding two pixels
Distance is also fixed, and the focal length of lens is known, and the relationship of similar triangles, it can obtains the three of camera respectively
Tie up position coordinates.
Figure 10 is pixel coordinate and image coordinate Transformation Graphs:
Pixel coordinate system and image coordinate system are all on imaging plane, and only respective origin and linear module are different.Image
The origin of coordinate system is the intersection point of camera optical axis and imaging plane, is the midpoint of imaging plane under normal conditions or is principal point.
The unit of image coordinate system is mm, belongs to physical unit.And the unit of pixel coordinate system is pixel, we describe one usually
Pixel is all several column of several rows.
The transformation relation of image coordinate system and pixel coordinate system is obtained from pixel coordinate and image coordinate, and then obtains pixel
The conversion of coordinate and image coordinate, acquire the subpoint of LED on the image sensor to center sensor point distance.
Attached drawing 11 is the building schematic diagram of the mobile vector based on optical flow method:
Camera is located on service robot, positions camera namely to robot localization.It is mounted on service robot
CMOS camera open image pickup mode, by variable length code technology emit with receive mentioned in based on visible light communication
Imaging location technology detects the position of LED lamp in initial image frame and determines the initial position of service robot.One
As in the case of, detect LED lamp by the image pixel intensities in image.However when CMOS camera is moved relative to LED lamp
When, LED lamp can generate blurring effect due to movement on the image, and due to the interference of environment light.So as to cause LED pixel
It selects and is mixed with environment light image vegetarian refreshments, to be difficult to differentiate.Geometrical characteristic and color coding of the application for LED lamp
The characteristics of, according to image pixel intensities, the light stream of LED lamp and the statistical information of previous frame image on imaging sensor, pass through
Bayesian criterion determines posterior probability, and the signal optical source of relative motion in video is quickly identified and locked, then obtains LED light
Have the ID location information of coding, the final real-time quick positioning realized to Information Mobile Service robot.
1, image pixel intensities
Under normal circumstances, visual optic communication is all based on institute's captured image image pixel intensities to distinguish LED signal point and back
It is general to calculate the condition that a certain pixel is LED signal point by obtaining the image pixel intensities of image in this application for sight spot pixel
Rate, it is assumed that the intensity of the ith pixel point of n-th frame image is f1 (n), wherein f1 (n)It is considered to have Gaussian Profile.
2, optical flow method
LED pixel moving direction light stream figure in image can be constructed by the continuous frame image of two width, and then obtains server
Device people moving direction since the time interval of adjacent two field pictures is very short, it can be considered that on two inter frame images LED pixel
Intensity is almost consistent, and Figure 11 illustrates how the motion vector of construction pixel, by the position of present frame LED pixel point with
Previous frame corresponds to the locality of LED pixel point and the variation of amplitude, it can determines the moving condition of pixel, wherein have
The movement pixel of effect passes through the threshold value in setting position direction and amplitude function.When lower than threshold value, it is believed that be stable；Higher than threshold
Value, just effectively to move pixel, calculates in image after the effective mobile vector of each LED pixel, can be obtained
The motion state of service robot.
3, bayesian criterion
Using the feature of image produced by optical flow method and image pixel intensities as class conditional probability density, come further according to Bayesian formula
Calculate the posterior probability that each pixel in image belongs to background dot or LED signal point.By the pixel on every frame image point
For two classes: W1Class, that is, LED lamp and W2Class, that is, background pixel.Classified with probability function P ∈ [0,1].It will be by above-mentioned light stream
θ, r and f of method and the collected pixel of image pixel intensities1 (n)All it is recorded as in f1 (n), as class conditional probability density letter
Number.And in the (n-1)th frame image, the number of LED pixel point is divided by the number probability obtained of total pixel as priori
Probability.For n-th frame, W1Class and W2The posterior probability of class is by Bayesian formula by class conditional probability density and the (n-1)th frame image
Prior probability calculate, judge that each pixel belongs to background dot or LED in n-th frame image by Bayes rule
Signaling point.
In traditional visual pursuit algorithm, since the movement of camera will lead to the picture of LED signal point in captured images
Plain intensity fogs and mixes with environment point, it is difficult to realize the tracking to LED lamp.The method that the application is proposed can
To reduce pixel intensity to the contribution degree of tracing and positioning, by θ, r and f1 (n)Three determines class conditional probability density simultaneously, makes
It obtains in camera high-speed motion, can also accurately track LED lamp.
Figure 12 is the visible light communication indoor locating system structural schematic diagram based on ROS service robot
The application sets up a indoor locating system applied on service robot platform, i.e., based on ROS service robot
Visible light communication indoor locating system.Entire robot positioning system includes three subsystems:
System 1: the visible light communication positioning system of view-based access control model perception；
System 2: the SLAM positioning system based on Cartographer；
System 3: the inertial positioning system based on inertial sensor.
Attached drawing 13 is the visible light communication positioning system of view-based access control model perception
System 1: the visible light communication positioning system of view-based access control model perception
The application uses CMOS camera as visible light communication positioning system receiving end optical signal detecting device, detailed process
It is:
Step 1, in static or initialization, capture single-frame images, the presence of the lamps and lanterns in detection image using CMOS camera；
Step 2, the LED identification characteristics for decoding its transmission；
Step 3, by capturing the position ID by mutilation long color coding techniques coding using the roller shutter effect of CMOS camera
Confidence breath；
Location technology is imaged by single camera to determine the three-dimensional location coordinates of service robot in step 4；
Step 5, service robot on the move, by the image capture frequency that setting is fixed, using being proposed based on shellfish
The visual pursuit algorithm of this criterion of leaf realizes the continuous capture and processing of image, can be carried out continuously the update of location information, from
And realize positioning in real time.
Figure 14 is the service robot visible light communication positioning system schematic illustration in true environment:
For product when actually coming into operation, the number of LED light is unlike only several as shown in Figure 13, but such as Figure 14
It is shown often to there is the LED light of huge number in practical architectural environment, in the practical, commercial of positioning system, lay,
The links such as operation and maintenance must be easy to use.Therefore the calculation of three kinds of ID that fast coding LED lamp position is set a kind of is designed
Method so that the ID of LED light include 1, floor ID1,2, room number ID2 and 3, LED room relative position ID3.
Attached drawing 15 is service robot positioning system flow chart
Self poisoning precision Shortcomings are caused in order to overcome service robot to rely solely on visible light communication location technology, example
Such as, there are the situations that LED is damaged in some room.The application merges location technology with visible light using SLAM.
Cartographer is Google in the library 2D and 3DSLAM that the ROS system that on October 06th, 2016 increases income is supported.
SLAM algorithm, which combines, comes from multiple sensors, for example, the data of LIDAR, IMU and camera, the synchronous position for calculating sensor
And draw the environment around sensor.In combination with the visible light communication location technology and Cartographer of view-based access control model perception
Open source two kinds of the library SLAM technology can carry out positioning coordinates correction in real time to service robot and real-time positioning map creates, raising
The stability of entire positioning system, especially solving indoor environment accidental power failure, LED light accidental damage etc. causes visible light logical
Believe the harsh conditions of positioning system cisco unity malfunction, the SLAM is Simultaneous Localization And
Mapping, instant positioning and map structuring.
Robot realizes that its service function is the pith that the application considers.Traditional robot is in navigator fix technology
Aspect all uses a variety of inertial sensors, such as accelerometer and gyroscope.In practical application scene, no beacon is considered
Environment, that is, the environment that LED lamp is not present or does not work normally.Inertial Measurement Unit is used under no beaconing environment
The data such as acceleration, angular speed, magnetic force and pressure of the IMU aware services robot during traveling, and utilize these data
The reckoning of step-length and direction is carried out to service robot, thus achieve the purpose that carry out locating and tracking to service robot, wherein
The process related generally to has gait detection, step-length and direction calculating.The positioning knot of the comprehensive visible light communication positioning system of the application
Fruit has the positioning result of beacon and inertial positioning system to determine their error according to actual test result that is, without beacon
Value carries out final positioning coordinate using particle filter accordingly and estimates, and then reduces the shadow caused by accumulating because of position error
It rings.Single positioning system poor compatibility, the defect of adaptive capacity to environment difference in practical applications are avoided simultaneously.SLAM positioning system
System utilizes updated positioning Coordinate Adjusting map, advanced optimizes the navigation effect of service robot, the whole work of system
It is as shown in figure 15 to make process.The english abbreviation of IMU, that is, Inertial Measurement Unit.
The application uses the long color coding techniques of mutilation.Traditional balance coding method mB/nB(m < n) word length be fixed
, it is unified to be transmitted with nB bit length, sacrifice a part of efficiency of transmission.If include in the binary string of mB bit length 0 and 1
Number is close, and variable length code technology still transmits this binary string with mB bit length.Traditional communication is needed before burst of data
Byte-header is added as start bit and check bit.It is encoded by the color for using lamps and lanterns different base colors as color coded word section header
Technology uses the less color head of occupy-place, improves code efficiency, and the low influence of code efficiency for improving imaging positioning system is fixed
The problem of position real-time.
The application proposition weakens the problem of pixel saturation using the method for second order polynomial fit.By to capturing
To image in each row carry out second order polynomial fit and select a suitable gray value, gray scale selected by all rows
Value is composed a column matrix and uses instead of initial data as decoding, can avoid effect of bursting forth, and then it is logical to solve vision
Blooming in letter and influence communication decoding the problem of.
The application proposes a kind of video frequency object tracking technology based on bayesian criterion, by taking to CMOS camera
Image sequence analyzed, the region where LED lamp is detected in initial image frame, according to the picture on imaging sensor
Plain intensity, optical flow method and bayesian criterion dynamically estimate the position of LED lamp in subsequent each frame, and then are increased to
The real-time of positioning, solves the problems, such as the real-time of service robot tracking and positioning.
Service robot positioning system detailed process is as follows:
Indoor service robot based on VLC carries out indoor positioning using visible light communication technology, installs in robot following
Three kinds of sensors, i.e. CMOS camera, kinect sensor and inertial sensor, write drive respectively on these three sensors
The circuit of dynamic program, and the positioning system with indoor service robot is positioned respectively
CMOS camera working principle is: CMOS camera installs start-up circuit, and obtains compressed data using opencv, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and synchronizes and is decoded, later, utilizes video
Driving, is then decoded image, is detected by LED light, obtain the ID number of LED light, so that the position for obtaining robot is sat
Mark, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Kinect sensor working principle is: Kinect sensor installs start-up circuit, and is obtained using cartographer
Compressed data, then compressed data is transmitted to by network or WIFI the transport layer of positioning system, and synchronize and be decoded, it
Afterwards, using video drive, then image is decoded, by feature extraction, instant map is created, to obtain robot
Position coordinates, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Inertial sensor working principle is: inertial sensor installs start-up circuit, and obtains metrical information using Kalman, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and signal processing is carried out to metrical information, it
Afterwards, posture information is updated, then by reckoning, the estimation of the position coordinates of robot is carried out based on particle filter algorithm.
Claims (5)
1. a kind of indoor service robot based on VLC, it is characterized in that: the indoor service robot based on VLC uses visible light
The communication technology carries out indoor positioning, installs following three kinds of sensors in robot, i.e., CMOS camera, kinect sensor with
And inertial sensor, write the circuit of driver respectively on these three sensors, and respectively with indoor service robot
Positioning system positioned
CMOS camera working principle is: CMOS camera installs start-up circuit, and obtains compressed data using opencv, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and synchronizes and is decoded, later, utilizes video
Driving, is then decoded image, is detected by LED light, obtain the ID number of LED light, so that the position for obtaining robot is sat
Mark, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Kinect sensor working principle is: Kinect sensor installs start-up circuit, and is obtained using cartographer
Compressed data, then compressed data is transmitted to by network or WIFI the transport layer of positioning system, and synchronize and be decoded, it
Afterwards, using video drive, then image is decoded, by feature extraction, instant map is created, to obtain robot
Position coordinates, then calculated by quantitative target, the estimation of the position coordinates of robot is carried out based on particle filter algorithm；
Inertial sensor working principle is: inertial sensor installs start-up circuit, and obtains metrical information using Kalman, then
Compressed data is transmitted to the transport layer of positioning system by network or WIFI, and signal processing is carried out to metrical information, it
Afterwards, posture information is updated, then by reckoning, the estimation of the position coordinates of robot is carried out based on particle filter algorithm.
2. the indoor service robot according to claim 1 based on VLC, it is characterized in that: obtaining machine by following steps
Device people's product:
Step 1: carrying out theoretical calculation first with VLC and visual identity and carry out digital modeling using computer modeling software；
Step 2: be respectively to obtain the numerical chracter of visible light communication using following three kinds of technologies:
2.1: utilizing the imaging mechanism of CMOS camera；
2.2: utilizing the mB/nB variable length code technology of VLC；
2.3: utilizing the imaging Location Mechanism of single camera；
Step 3: by optical flow method and bayesian criterion, realizing the tracking to LED lamp track；
Step 4: the big positioning system of fusion following two, and VLC movable object tracking positioning system is developed, two big positioning system difference
It is:
4.1:cartographer open source SLAM positioning system；
4.2: inertial positioning system；
Step 5: obtaining high-accuracy position system in the service robot room based on visible light communication；
Step 6: positioning performance test and stability test；
Step 7: final to obtain service robot product.
3. the indoor service robot according to claim 1 based on VLC, it is characterized in that: the acquisition LED lamp institute
The ID location information of transmission, basic procedure is:
The image of intake is obtained a frame image of caching by step 1, camera；
Step 2 pre-processes the image of caching；
Step 3 carries out downlight image segmentation to image；
Step 4, the extraction that downlight feature is carried out to image；
Step 5 does not include downlight, then gives up image, includes downlight, into next step；
Step 6, the selection into striate region；
Step 7 pre-processes striate region；
Whether step 8, analysis striped include that two color heads are given up if do not included, if comprising into next step；
Step 9, carry out color verification, color verification mainly color difference is verified, in RGB equipment, due to display screen with
The difference of camera lens manufacturing process and external visible light interference, cause the identical color of binary data, by showing, shooting with video-corder
After technique, color difference will result in, still, color difference is acceptable to a certain extent；
Step 10, by analysis, sees whether the data of transmission are correct, if incorrect, give up, if correctly, executing solution
Code.
4. the indoor service robot according to claim 1 based on VLC, it is characterized in that: the positioning system is determined
Position the following steps are included:
Step 1, camera are located on service robot, to camera positioning namely to robot localization,
The CMOS camera being mounted on service robot opens image pickup mode, in the transmitting of variable length code technology and receiving
Mentioned detects the position of LED lamp and determines to take in initial image frame based on visible light communication imaging location technology
It is engaged in the initial position of robot, however when CMOS camera is moved relative to LED lamp, LED lamp on the image can be due to
Movement generates blurring effect, and due to the interference of environment light, selects so as to cause LED pixel and be blended in one with environment light image vegetarian refreshments
It rises, therefore enters step 2；
Step 2, visual optic communication are all based on institute's captured image image pixel intensities to distinguish LED signal point and background dot picture
Element calculates the conditional probability that a certain pixel is LED signal point by obtaining the image pixel intensities of image, it is assumed that n-th frame image
Ith pixel point intensity be f1 (n), wherein f1 (n)With Gaussian Profile；
Step 3 goes out LED pixel moving direction light stream figure in image, and then the service of obtaining by the continuous frame picture construction of two width
Robot moving direction is since the time interval of adjacent two field pictures is very short, and the image pixel intensities of LED are several on two inter frame images
It is consistent, passes through the locality of the position of present frame LED pixel point LED pixel point corresponding with previous frame and the change of amplitude
Change, it can determine the moving condition of pixel, wherein effective movement pixel passes through setting position direction and amplitude function
Threshold value, when be lower than threshold value, it is believed that be stable；Higher than threshold value, just effectively to move pixel, calculate each in image
After the effective mobile vector of LED pixel, the motion state of service robot can be obtained, this step is optical flow method；
Step 4, using the feature of image produced by step 2 and step 3 as class conditional probability density, further according to Bayesian formula
The posterior probability that each pixel in image belongs to background dot or LED signal point is calculated, by the pixel on every frame image
It is divided into two classes: W1Class, that is, LED lamp and W2Class, that is, background pixel is classified with probability function P ∈ [0,1], will be by above-mentioned light
θ, r and f of stream method and the collected pixel of image pixel intensities1 (n)All it is recorded as in f1 (n), as class conditional probability density
Function, and in the (n-1)th frame image, the number of LED pixel point is divided by the number probability obtained of total pixel as first
Probability is tested, for n-th frame, W1Class and W2The posterior probability of class is by Bayesian formula by class conditional probability density and the (n-1)th frame figure
The prior probability of picture calculates, and judges that each pixel belongs to background dot still in n-th frame image by Bayes rule
LED signal point.
5. the indoor service robot according to claim 2 based on VLC is compiled it is characterized in that: the mB/nB mutilation is long
Code technical solution be: former word length be m symbol, if comprising 0 and 1 several difference be greater than 2, from word length be n code
Member concentrate select a symbol so that comprising 0 and 1 several difference be not more than 2, substitute original symbol and transmitted；It is former
Word length be m symbol, if comprising 0 and 1 several difference be not more than 2, still transmitted with m symbols, mutilation
The purpose of long codes technology is to obtain higher illumination stability simultaneously, improves code efficiency,
It comprises the concrete steps that:
Step 1, initial data pass through the long color coding of mB/nB mutilation；
Step 2, driving LED light, to obtain specified color information value；
Step 3, camera obtain its color information value；
Step 4 passes through decoding, obtains new color data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910702032.9A CN110296707A (en) | 2019-07-31 | 2019-07-31 | Indoor service robot based on VLC technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910702032.9A CN110296707A (en) | 2019-07-31 | 2019-07-31 | Indoor service robot based on VLC technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110296707A true CN110296707A (en) | 2019-10-01 |
Family
ID=68032415
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910702032.9A Withdrawn CN110296707A (en) | 2019-07-31 | 2019-07-31 | Indoor service robot based on VLC technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110296707A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111062337A (en) * | 2019-12-19 | 2020-04-24 | 北京迈格威科技有限公司 | People flow direction detection method and device, storage medium and electronic equipment |
| CN111928852A (en) * | 2020-07-23 | 2020-11-13 | 武汉理工大学 | Indoor robot positioning method and system based on LED position coding |
| CN112082556A (en) * | 2020-09-08 | 2020-12-15 | 杭州电子科技大学 | An indoor positioning method based on color ring light landmarks |
| CN112229398A (en) * | 2020-09-03 | 2021-01-15 | 宁波诺丁汉大学 | Navigation system and method for indoor fire escape |
| CN113324550A (en) * | 2021-06-30 | 2021-08-31 | 维飞科技有限公司 | Flight path correction method based on improved flight path integral and storage medium |
| CN113447023A (en) * | 2021-06-18 | 2021-09-28 | 上海宏景智驾信息科技有限公司 | Indoor visual positioning method based on ceiling lamp |
| CN115118338A (en) * | 2022-05-31 | 2022-09-27 | 重庆工程职业技术学院 | A real-time positioning system for remote tunnel construction personnel |
| CN115507854A (en) * | 2022-09-22 | 2022-12-23 | 广东省智能机器人研究院 | Mobile robot vision positioning method in indoor weak texture environment |
-
2019
- 2019-07-31 CN CN201910702032.9A patent/CN110296707A/en not_active Withdrawn
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111062337B (en) * | 2019-12-19 | 2023-08-04 | 北京迈格威科技有限公司 | People stream direction detection method and device, storage medium and electronic equipment |
| CN111062337A (en) * | 2019-12-19 | 2020-04-24 | 北京迈格威科技有限公司 | People flow direction detection method and device, storage medium and electronic equipment |
| CN111928852B (en) * | 2020-07-23 | 2022-08-23 | 武汉理工大学 | Indoor robot positioning method and system based on LED position coding |
| CN111928852A (en) * | 2020-07-23 | 2020-11-13 | 武汉理工大学 | Indoor robot positioning method and system based on LED position coding |
| CN112229398A (en) * | 2020-09-03 | 2021-01-15 | 宁波诺丁汉大学 | Navigation system and method for indoor fire escape |
| CN112082556A (en) * | 2020-09-08 | 2020-12-15 | 杭州电子科技大学 | An indoor positioning method based on color ring light landmarks |
| CN112082556B (en) * | 2020-09-08 | 2024-04-30 | 杭州电子科技大学 | Indoor positioning method based on color ring lamp landmark |
| CN113447023A (en) * | 2021-06-18 | 2021-09-28 | 上海宏景智驾信息科技有限公司 | Indoor visual positioning method based on ceiling lamp |
| CN113447023B (en) * | 2021-06-18 | 2022-11-01 | 上海宏景智驾信息科技有限公司 | Indoor visual positioning method based on ceiling lamp |
| CN113324550A (en) * | 2021-06-30 | 2021-08-31 | 维飞科技有限公司 | Flight path correction method based on improved flight path integral and storage medium |
| CN113324550B (en) * | 2021-06-30 | 2024-03-26 | 维飞科技有限公司 | Track correction method based on improved track integration and storage medium |
| CN115118338A (en) * | 2022-05-31 | 2022-09-27 | 重庆工程职业技术学院 | A real-time positioning system for remote tunnel construction personnel |
| CN115118338B (en) * | 2022-05-31 | 2023-07-25 | 重庆工程职业技术学院 | A real-time positioning system for remote tunnel construction personnel |
| CN115507854A (en) * | 2022-09-22 | 2022-12-23 | 广东省智能机器人研究院 | Mobile robot vision positioning method in indoor weak texture environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110296707A (en) | Indoor service robot based on VLC technology | |
| US11328158B2 (en) | Visual-inertial positional awareness for autonomous and non-autonomous tracking | |
| CN105930819B (en) | Real-time city traffic lamp identifying system based on monocular vision and GPS integrated navigation system | |
| US20230236280A1 (en) | Method and system for positioning indoor autonomous mobile robot | |
| Naimark et al. | Circular data matrix fiducial system and robust image processing for a wearable vision-inertial self-tracker | |
| US7231063B2 (en) | Fiducial detection system | |
| CN111753757B (en) | Image recognition processing method and device | |
| CN109405824A (en) | A kind of multi-source perceptual positioning system suitable for intelligent network connection automobile | |
| CN109341694A (en) | An autonomous positioning and navigation method for a mobile detection robot | |
| Xing et al. | Mapping human activity volumes through remote sensing imagery | |
| CN108986158A (en) | A kind of across the scene method for tracing identified again based on target and device and Computer Vision Platform | |
| WO2020083103A1 (en) | Vehicle positioning method based on deep neural network image recognition | |
| CN105608417A (en) | Traffic signal lamp detection method and device | |
| CN112233079B (en) | Method and system for fusing images of multiple sensors | |
| CN110458089A (en) | A kind of naval target interconnected system and method based on the observation of height rail optical satellite | |
| CN114252883A (en) | Target detection method, apparatus, computer device and medium | |
| CN115235478A (en) | Intelligent vehicle localization method and system based on visual label and laser SLAM | |
| CN113870317B (en) | Visible light positioning method and system based on enhanced visual target tracking | |
| CN111505577A (en) | Mobile vehicle localization method based on visible light communication | |
| Wu et al. | Multimodal collaboration networks for geospatial vehicle detection in dense, occluded, and large-scale events | |
| CN107196704A (en) | Method of locating terminal, device and electronic equipment | |
| Kampmann et al. | Investigating Outdoor Recognition Performance of Infrared Beacons for Infrastructure-based Localization | |
| CN115755985A (en) | Unmanned aerial vehicle landing method based on image processing | |
| CN114299420A (en) | Positioning method and device based on screen optical communication and server | |
| Liu et al. | Dynamic visible light positioning based on enhanced visual target tracking |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20191001 |