CN104848851B - Intelligent Mobile Robot and its method based on Fusion composition - Google Patents

Intelligent Mobile Robot and its method based on Fusion composition Download PDF

Info

Publication number
CN104848851B
CN104848851B CN201510289939.9A CN201510289939A CN104848851B CN 104848851 B CN104848851 B CN 104848851B CN 201510289939 A CN201510289939 A CN 201510289939A CN 104848851 B CN104848851 B CN 104848851B
Authority
CN
China
Prior art keywords
lt
gt
mo
mi
mrow
Prior art date
Application number
CN201510289939.9A
Other languages
Chinese (zh)
Other versions
CN104848851A (en
Inventor
刘加科
孔令文
田晓璐
韩磊
付崇光
秦振华
孙凯
Original Assignee
山东鲁能智能技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 山东鲁能智能技术有限公司 filed Critical 山东鲁能智能技术有限公司
Priority to CN201510289939.9A priority Critical patent/CN104848851B/en
Publication of CN104848851A publication Critical patent/CN104848851A/en
Application granted granted Critical
Publication of CN104848851B publication Critical patent/CN104848851B/en

Links

Abstract

The invention discloses a kind of Intelligent Mobile Robot based on Fusion composition and its method, including robot moving platform, there are industrial computer, odometer, tachogenerator, vision collecting device and laser radar on the robot moving platform;The industrial computer is connected respectively with climb displacement device, vision collecting device and laser radar;Climb displacement device is connected with tachogenerator.Beneficial effect of the present invention:Visual pattern processing positioning of the present invention, laser radar synchronous superposition are two kinds of decision positioning methods, climb displacement device reckoning is a kind of relative positioning mode, it is successfully realized collective effect of two kinds of absolute fix modes in composition, relative positioning mode is corrected by absolute fix mode, the continuous correction of climb displacement device is realized.

Description

Intelligent Mobile Robot and its method based on Fusion composition

Technical field

The present invention is under the jurisdiction of localization for Mobile Robot and field of navigation technology, more particularly to it is a kind of based on odometer, vision, The Intelligent Mobile Robot and method of the multisensor Data Fusion technology composition such as laser radar.

Background technology

With Intelligent Mobile Robot in recent years deeper into application, complicated substation and power transformation patrol task Navigation mode and navigation and positioning accuracy for crusing robot propose higher requirement, in order to preferably adapt to complicated Substation, the need for meeting complicated substation inspection task, increasing airmanship input is tested and put into should With laser navigation technology is one of them.In robotic laser navigation procedure, realize that autonomous positioning and high accuracy navigation are to patrol The important prerequisite that robot completes patrol task is examined, and realizes positioning and the key of high accuracy navigation and is to set up complete and accurate Substation map, therefore, transformer station's map structuring for Intelligent Mobile Robot carry out laser navigation have it is important Meaning.The method for substation map structuring mainly has following several at present:

1st, substation map structuring is carried out using 3D (three-dimensional) laser scanner.This method is by transformer station The interval multiple positions of selection carry out whole station multidrop environment data acquisition.In each position, 3D laser scanners carry out level 360 ° of three-dimensional environment data acquisitions.After whole station multidrop environment data acquisition terminates, use environment data processing software is to collecting Multidrop environment data carry out splicing reproduction processes, it is established that transformer station's three-dimensional environment map.On the basis of three-dimensional environment map It is upper to be cut, cut height and make every effort to identical with the height of Intelligent Mobile Robot laser radar, the Final finishing output of cutting The available two-dimentional substation map of Intelligent Mobile Robot location navigation.Although this method builds figure precision height, and environment is again Existing effect is good, but this method cost is high, needs to be cut repeatedly for the uneven region of local transformer station's physical features, and right It is difficult in the modification of local map.

2nd, using the method for carrying out transformer station's map structuring based on odometer, gyroscope, laser radar.This method is machine People is roamed in whole station, and carrying out environmental data by the odometer, gyroscope, laser radar that are installed on robot body adopts Collection.After data environment data acquisition terminates, data are entered into row format and change edlin of going forward side by side, final output two dimension substation Map.Although this method scheme has certain feasibility, environmental data collecting, location navigation are realized in same robot It is upper integrated, but the program is belonged to relative positioning mode and there is error, environmental data collecting process due to odometer and gyroscope Middle error cannot be corrected always, can have larger accumulated error for a long time or after large scale reckoning, annular is returned Road can not be closed, and this is accomplished by the processing of later stage environmental data, in addition to Data Format Transform and editor, in order to correct accumulation The map distortion that error band comes, in addition it is also necessary to introduce complicated ICP (closest approach loop iteration matching algorithm), spring model etc. and calculate Method, scheme is complicated, and a large amount of careful map modification adjustment work are also carried out in addition, and normative reference is obscured, and workload is big and deposits The problems such as accuracy of map after modification is not high, reproduction effects are bad.

The content of the invention

The purpose of the present invention is exactly to solve the above problems, it is proposed that one kind is based on multisensor Data Fusion technology structure The Intelligent Mobile Robot and its method of figure, by constantly carrying out feedback compensation to odometer, effectively reduce odometer For a long time or large scale reckoning cumulative errors, solve the loop checking installation that odometer cumulative errors cause and do not close, map The problems such as deformation.

To achieve these goals, the present invention is adopted the following technical scheme that:

A kind of Intelligent Mobile Robot based on Fusion composition, including robot moving platform, institute Stating on robot moving platform has industrial computer, odometer, tachogenerator, vision collecting device and laser radar;

Industrial computer:It is carried to the computer on robot moving platform;

Climb displacement device:Receive tachogenerator signal, continuous output device people current positional information and course angle Information, and described information is sent to industrial computer by communication line;

Vision collecting device:For gathering the mark line image on Intelligent Mobile Robot Roam Path, and by logical Letter circuit is sent to industrial computer;

Laser radar:For realizing the scanning to Intelligent Mobile Robot operation area environmental data, and by scanning figure As being sent to industrial computer by communication line;

The industrial computer is connected respectively with climb displacement device, vision collecting device and laser radar;Climb displacement device It is connected with tachogenerator.

The vision collecting device includes video camera, and camera lens are parallel to the ground, is distributed around camera lens There is polishing LED light array, video camera is connected by netting twine with industrial computer.

On the Intelligent Mobile Robot Roam Path of layout area, tag line is set, the tag line is that color is bright The aobvious colored colour band or line for being different from road surface color.

A kind of patterning process of the Intelligent Mobile Robot based on Fusion composition, including following step Suddenly:

(1) before Intelligent Mobile Robot patterning process starts, the substation inspection machine in target pattern region Tag line is set on people's Roam Path;

(2) set up directly as origin (0,0,0) using the position that Intelligent Mobile Robot climb displacement device is started working Angular coordinate system, climb displacement device is patrolled by receiving progress reckoning to tachogenerator signal, and exporting t transformer station Examine the current positional information of robot and course angle information;

Meanwhile, the mark line image on vision collecting device collection Intelligent Mobile Robot mobile platform Roam Path, Handled by visual pattern and identify course angle information of the t Intelligent Mobile Robot mobile platform relative to tag line;

Environment around Laser Radar Scanning robot moving platform Roam Path, obtains different laser radar datas, The data include the distance of sampling anglec of rotation pip corresponding with the angle, and scan data is transmitted to work by Ethernet Control machine;

(3) environmental data that t is collected according to laser radar uploads the positional information of t with climb displacement device Local composition is carried out to scanning area with course angle information, and the position for exporting Intelligent Mobile Robot by synchronous positioning is believed Breath and course angle information;

(4) in the course angle information for the Intelligent Mobile Robot for obtaining synchronous localization process and step (2) by regarding Feel that the course angle information that image procossing is identified is weighted filtering process, the course angle letter for the Du Genggao that established trust by filtering Breath;

(5) positional information obtained synchronous localization process and weighted filtering processing after course angle feedback of the information to go Journey computing device realizes the correction to climb displacement device;

(6) repeated the above steps (2)~(5), and local composition is simultaneously carried out the global composition of fusion output by loop iteration.

In the step (2), climb displacement device to tachogenerator signal by receiving the method for carrying out reckoning For:

Wherein, Sr (t-1), Sl (t-1) are respectively robot moving platform right wheel and revolver when the t-1 moment is to t Between be spaced the distance passed by, d is robot moving platform wheelspan;[X (t), Y (t), W (t)] is t robot location's information With course angle information.

In the step (2), handled by visual pattern and identify that t Intelligent Mobile Robot mobile platform is relative It is in the specific method of the course angle information of tag line::

(2-1) image calibration:The calibrating parameters obtained using the camera calibration stage are demarcated to each two field picture, to disappear The pattern distortion brought except camera lens;

(2-2) carries out color model transformation to chromatic image, is HSI model images RGB model conversions;

(2-3) target image is split, and selects region-of-interest in HSI model images, passes through the H and S of determination threshold value pair Image is split, and extracts Characteristic Contrast image;

(2-4) Morphological scale-space, is measured and is extracted to characteristics of image by picture structure element, by corrosion and it is swollen Swollen morphological method is to image procossing, to facilitate the identification and analysis to feature;

(2-5) extracts target signature, determines often row target signature center point coordinate by rim detection, passes through particle analysis The angle of tag line and image vertical central axis line is calculated, course angle information w of the robot in working region is global is determined (t) vision positioning, that is, is realized.

Setting identification line image processing time interval need to be shifted to an earlier date for visual pattern processing, vision collecting device is every setting Time interval is performed once, is handled and obtained and output device people course angle information by visual pattern.

The specific method of the step (3) is:

3-1) the point set data that laser radar is collected are transformed under rectangular coordinate system;

Clustering processing 3-2) is carried out to the point set under rectangular coordinate system using clustering distance threshold value index;

3-3) point set after cluster is carried out curve fitting using least square method, reference substance characteristic straight line side is obtained Journey, determines the center point coordinate of each reference substance characteristic straight line;

3-4) adjacent feature straight line is matched, by the local feature offset between each reference substance characteristic straight line Distribution average computation is carried out, the optimal offset of consecutive frame feature is drawn;

Previous frame data 3-5) are added into a little optimal offset obtained in the previous step, local map is obtained;Repeat Above-mentioned processing procedure;

3-6) by loop iteration, new laser data point set is matched with legacy data, synchronized update positional information and Course angle information.

The step 3-2) specific method be:

The distance between point set consecutive points are calculated by order, judged whether within clustering distance threshold value indication range, Consecutive points are clustered if in the range of;The isolated point beyond each cluster areas is will be independent of to remove.

The step 3-4) specific method be:

Front and rear adjacent each reference substance characteristic straight line central point distance of two frames is compared with characteristic distance threshold value, if adjacent spy Levy straight central point coordinates and be then considered the adjacent same reference substance characteristic straight line of two frames less than characteristic distance threshold value;Joined with a later frame The center point coordinate for examining thing characteristic straight line subtracts former frame correspondence reference substance feature point coordinates, obtains local feature offset, will Local feature offset between each reference substance characteristic straight line of consecutive frame carries out distribution average computation, show that consecutive frame feature is optimal Offset.

The beneficial effects of the invention are as follows:

Visual pattern processing positioning of the present invention, laser radar synchronous superposition are two kinds of decision positioning methods, Climb displacement device reckoning is a kind of relative positioning mode, is successfully realized two kinds of absolute fix mode being total in composition Same-action, relative positioning mode is corrected by absolute fix mode, realizes the continuous correction of climb displacement device.

The present invention effectively reduces climb displacement device in length by continuously carrying out feedback compensation to climb displacement device Time or large scale reckoning cumulative errors, solve the loop checking installation that climb displacement device cumulative errors cause and do not close, The problems such as map deformation.

The present invention is that scheme is simple, and cost is low, and feasibility is good.It is distributed with around visual image acquisition device camera lens Polishing LED, solves the strong and weak influence to visual pattern processing of illumination, is adapted to night-environment and carries out map structuring.

Brief description of the drawings

Fig. 1 is the Intelligent Mobile Robot structural representation of the invention based on multisensor Data Fusion technology composition;

Fig. 2 is the Intelligent Mobile Robot flow chart of work methods of the invention based on multi-technical fusion composition;

Wherein, 1, robot moving platform, 2, industrial computer, 3, odometer, 4, vision collecting device, 5, laser radar, 6, Driving wheel, 7, tag line.

Embodiment:

The present invention will be further described with embodiment below in conjunction with the accompanying drawings:

In Fig. 1, based on the Intelligent Mobile Robot of multisensor Data Fusion technology composition, it is moved including robot Platform 1, there is an industrial computer 2 on the platform, odometer 3, vision collecting device 4, laser radar 5, the driving wheel 6 of mobile platform, There is tag line 7 on ground.

Described robot moving platform 1, the embodiment is wheel moving platform form, if being characterized in, mobile platform has Dry driving wheel 6;

Described industrial computer 2, is characterized in it being the computer being carried on robot moving platform;

Described odometer 3, is characterized in receiving tachogenerator signal, and be connected with industrial computer by communication line;

Described vision collecting device 4, is characterized in, the device is installed on robot moving platform, possesses core and regard Feel acquisition elements video camera, camera lens are parallel to the ground, and video camera is connected by netting twine with industrial computer 2, in the embodiment Mako G-032 industrial cameras are used, the image vertical central axis line of video camera is parallel with tag line 7, in camera lens week Enclose and polishing LED array is distributed with;

The laser radar 5, is characterized in being installed on the front end or rear end of robot moving platform, and laser radar passes through logical Letter line is connected with industrial computer 2;

Described tag line 7, its feature is temporarily arranged on layout area robot Roam Path before composition progress, is marked Knowing line has the course angle information global relative to layout area determined;

Below in conjunction with the accompanying drawings 2 pairs the present invention relates to the substation inspection based on multisensor Data Fusion technology composition The method and step of robot is described further.

[1] in target pattern region, before robot patterning process starts, in layout area robot Roam Path On be provided with tag line, the tag line has the course information global relative to layout area determined.In the present embodiment, mark Line is the colored colour band or line that a kind of color is clearly distinguishable from road surface.

[2] rectangular coordinate system, start machine are set up using the position of robot odometer start-up operation as origin (0,0,0) People's system, starts laser radar, starts composition processing routine.The front of robot is x-axis direction in the embodiment, with x-axis It is vertical and with axle into the direction of right-handed coordinate system be y-axis.

[3] odometer carries out reckoning by being received to tachogenerator signal, and exports t substation inspection machine The current positional information of device people and course angle information, this is a kind of relative positioning mode;

Odometer by tachogenerator signal receive carry out reckoning method be:

Wherein, Sr (t-1), Sl (t-1) are respectively robot moving platform right wheel and revolver when the t-1 moment is to t Between be spaced the distance passed by, d is robot moving platform wheelspan;[X (t), Y (t), W (t)] is t robot location's information With course angle information.

Meanwhile, the mark line image on vision collecting device collection Intelligent Mobile Robot mobile platform Roam Path, Handled by visual pattern and identify course angle information of the t Intelligent Mobile Robot mobile platform relative to tag line, This is a kind of absolute fix mode;

Environment around Laser Radar Scanning robot moving platform Roam Path, obtains different laser radar datas, The data include the distance of sampling anglec of rotation pip corresponding with the angle, and scan data is transmitted to work by Ethernet Control machine;

Odometer reckoning, visual pattern processing, the scanning of laser radar environmental data, it is synchronous to carry out.In the embodiment, Odometer carries out reckoning by being received to tachogenerator signal, and constantly the positional information of output device people's t and Course angle information.Visual pattern processing module is to the mark line image on robot moving platform Roam Path in the embodiment Handled, and identify the angle of t tag line feature and image vertical central axis line, the angle is this robot at moment Course angle of the mobile platform relative to tag line.

Key step is as follows:

(1) image calibration.The calibrating parameters obtained using the camera calibration stage are demarcated to each two field picture, to eliminate The pattern distortion that camera lens are brought.

(2) color model transformation is carried out to chromatic image.What it is due to the system needs identification is coloured sign chromatape Or line, so needing to extract by the color model to interesting image.The color graphics processing space that the system is used Model is HSI models, and the model is changed by illumination condition is influenceed small, and H, which represents tone, S and represents saturation degree, I, represents brightness. RGB model conversions are HSI model images.

(3) target image is split.RIO (region-of-interest) is selected in HSI model images, by the H for learning stage determination Image is split with S threshold value, Characteristic Contrast image is extracted.

(4) Morphological scale-space.Characteristics of image is measured and extracted by picture structure element, by corroding and expanding Etc. morphological method to image procossing to facilitate the identification and analysis to feature.

(5) target's feature-extraction and parameter are calculated.The method that target's feature-extraction is used is to use 8 connected domains to enter image Row scanning.After feature extraction comes out, often row target signature center point coordinate is determined by rim detection, passes through particle analysis meter The angle of tag line and image vertical central axis line is calculated, because tag line is true in the global course information of robot work region It is fixed, this at moment robot relative to tag line course angle have calculated that just can determine that robot in working region it is global in Course information w (t), that is, realize vision positioning, and this is a kind of absolute fix mode.

It is computationally intensive because visual pattern processing module needs to take more resource, so needing to preset vision Image processing time interval, the time interval can be 5~10 times of odometer reckoning cycle, visual pattern processing module Each time interval is performed once, and by output device people course angle information w (t), vision collecting device is taken the photograph in the embodiment Polishing LED array is distributed with around camera, the strong and weak influence to visual pattern processing of illumination can be solved.Swashing in the embodiment Optical radar environmental data is scanned, and is that laser radar carries out rotation 190 degree or 270 degree or 360 degree of scannings, is obtained different laser thunders Up to data, the data include the distance of sampling anglec of rotation pip corresponding with the angle, and scan data is passed through into Ethernet Transmit to industrial computer.

[4] the composition module inside industrial computer uses SLAM (synchronous superposition method) method according to laser thunder Scanning area is patterned up to the environmental data collected.

The specific steps of SLAM methods:

(1) laser radar data coordinate system is changed.Under being polar coordinate system due to the point set data obtained from laser radar Data (d, w), in order to be handled under the same coordinate system, so needing to pass through (d*cosw, d*sinw) by laser radar point set data It is transformed under rectangular coordinate system.

(2) clustering processing is carried out to the point set under rectangular coordinate system using clustering distance threshold value index.Calculated by order The distance between adjacent point set, judges whether within cluster threshold range, is clustered consecutive points if in the range of. For being removed independently of the isolated point beyond each cluster areas.

(3) point set after cluster is carried out curve fitting.Using least square method to the point set march after cluster Line is fitted, and is obtained reference substance characteristic straight line equation, is determined the center point coordinate of each reference substance characteristic straight line.

(4) adjacent feature straight line is matched, by the local feature offset between each reference substance characteristic straight line Distribution average computation is carried out, the optimal offset of consecutive frame feature is drawn.Will front and rear adjacent each reference substance characteristic straight line center of two frames Point distance is compared with characteristic distance threshold value, and phase is considered if adjacent feature straight central point coordinates is less than characteristic distance threshold value The same reference substance characteristic straight line of adjacent two frames;Former frame correspondence reference is subtracted with the center point coordinate of a later frame reference substance characteristic straight line Thing feature point coordinates, obtains local feature offset, by the local feature offset between each reference substance characteristic straight line of consecutive frame Distribution average computation is carried out, the optimal offset of consecutive frame feature is drawn.

(5) by the offset added a little in previous step of previous frame data, local map is obtained;Repeat the above Process;

(6) by loop iteration, circulation pipe is by new laser data point set and old Data Matching, synchronized update position successively Confidence ceases and course angle information (Xt, Yt, Wt).

[5] the synchronous positioning position informations of SLAM and course angle information enter with visual pattern processing module positioning course angle information Row weighting is handled:

The formula of weighted filtering processing is as follows:

W'(t)=c*W (t)+(1-c) * w (t)+Pe

W (t) is by the synchronous course angle information for being positioned at the t robot that map structuring is obtained, w (t) in above formula The course angle information of the t robot obtained for visual pattern processing.C is weight coefficient, passes through the basis in debugging process Experience is set, and Pe is penalty coefficient because in climb displacement device and visual pattern processing procedure data exist it is asynchronous, A certain amount of penalty coefficient is added in updating formula, this penalty coefficient will be with reference to systematic sampling speed and calculating speed It is fixed, correct obtained course angle W'(t) complete one during the reckoning at odometer new cycle t+1 moment is input to The odometer correction in individual cycle.

[6] the coordinate information Wt ' after the positional information (Xt, Yt) after synchronous localization process and weighting processing is fed back The correction to odometer is realized to odometer.

[7] repeated the above steps [2]~[6], and local map is simultaneously carried out fusion output global map by loop iteration.

Although above-mentioned the embodiment of the present invention is described with reference to accompanying drawing, not to present invention protection model The limitation enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not Need to pay various modifications or deform still within protection scope of the present invention that creative work can make.

Claims (7)

1. a kind of patterning process of the Intelligent Mobile Robot based on Fusion composition, it is characterized in that, use Based on the Intelligent Mobile Robot of Fusion composition, including robot moving platform, the robot movement There are industrial computer, odometer, tachogenerator, vision collecting device and laser radar on platform;
Industrial computer:It is carried to the computer on robot moving platform;
Climb displacement device:Tachogenerator signal, the current positional information of continuous output device people and course angle information are received, And described information is sent to industrial computer by communication line;
Vision collecting device:For gathering the mark line image on Intelligent Mobile Robot Roam Path, and pass through order wire Road is sent to industrial computer;
Laser radar:Lead to for realizing the scanning to Intelligent Mobile Robot operation area environmental data, and by scan image Cross communication line and be sent to industrial computer;
The industrial computer is connected respectively with climb displacement device, vision collecting device and laser radar;Climb displacement device is with surveying Fast sensor connection;
Method comprises the following steps:
(1) before Intelligent Mobile Robot patterning process starts, the Intelligent Mobile Robot in target pattern region overflows Trip sets tag line on path;
(2) right angle is set up as origin (0,0,0) using the position that Intelligent Mobile Robot climb displacement device is started working to sit Mark system, climb displacement device carries out reckoning by being received to tachogenerator signal, and exports t substation inspection machine The current positional information of device people and course angle information;
Meanwhile, the mark line image on vision collecting device collection Intelligent Mobile Robot mobile platform Roam Path passes through Visual pattern processing identifies course angle information of the t Intelligent Mobile Robot mobile platform relative to tag line;
Environment around Laser Radar Scanning robot moving platform Roam Path, obtains different laser radar datas, the number According to the distance of sampling anglec of rotation pip corresponding with the angle is included, and scan data is transmitted to industry control by Ethernet Machine;
(3) environmental data that t is collected according to laser radar uploads the positional information and boat of t with climb displacement device Carry out local composition to scanning area to angle information, and by synchronous positioning export Intelligent Mobile Robot positional information and Course angle information;
(4) the course angle information for the Intelligent Mobile Robot for obtaining synchronous localization process in step (2) with passing through vision figure As the course angle information that processing is identified is weighted filtering process, by the course angle information for filtering the Du Genggao that establishes trust;
(5) the course angle feedback of the information after the positional information obtained synchronous localization process and weighted filtering processing is to stroke counter Calculate correction of the device realization to climb displacement device;
(6) repeated the above steps (2)~(5), and local composition is simultaneously carried out the global composition of fusion output by loop iteration.
2. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 1 Method, it is characterized in that, in the step (2), climb displacement device carries out reckoning by being received to tachogenerator signal Method is:
<mrow> <mi>X</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>X</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mfrac> <mrow> <mi>S</mi> <mi>r</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mi>S</mi> <mi>l</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> <mn>2</mn> </mfrac> <mo>*</mo> <mi>c</mi> <mi>o</mi> <mi>s</mi> <mrow> <mo>(</mo> <mi>W</mi> <mo>(</mo> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> <mo>)</mo> <mo>)</mo> </mrow> </mrow>
<mrow> <mi>Y</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>Y</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mfrac> <mrow> <mi>S</mi> <mi>r</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mi>S</mi> <mi>l</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> <mn>2</mn> </mfrac> <mo>*</mo> <mi>s</mi> <mi>i</mi> <mi>n</mi> <mrow> <mo>(</mo> <mi>W</mi> <mo>(</mo> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> <mo>)</mo> <mo>)</mo> </mrow> </mrow> 1
<mrow> <mi>W</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>W</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>+</mo> <mfrac> <mrow> <mi>S</mi> <mi>r</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>-</mo> <mi>S</mi> <mi>l</mi> <mrow> <mo>(</mo> <mi>t</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> <mi>d</mi> </mfrac> </mrow>
Wherein, Sr (t-1), Sl (t-1) be respectively robot moving platform right wheel and revolver at the t-1 moment between the t time Every the distance passed by, d is robot moving platform wheelspan;[X (t), Y (t), W (t)] is t robot location information and boat To angle information.
3. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 1 Method, it is characterized in that, in the step (2), handled by visual pattern and identify t Intelligent Mobile Robot mobile platform Specific method relative to the course angle information of tag line is::
(2-1) image calibration:The calibrating parameters obtained using the camera calibration stage are demarcated to each two field picture, to eliminate phase The pattern distortion that machine camera lens is brought;
(2-2) carries out color model transformation to chromatic image, is HSI model images RGB model conversions;
(2-3) target image is split, and region-of-interest is selected in HSI model images, by the H and S of determination threshold value to image Split, extract Characteristic Contrast image;
(2-4) Morphological scale-space, is measured and is extracted to characteristics of image by picture structure element, passes through what is corroded and expand Morphological method is to image procossing, to facilitate the identification and analysis to feature;
(2-5) extracts target signature, determines often row target signature center point coordinate by rim detection, is calculated by particle analysis Go out the angle of tag line and image vertical central axis line, determine course angle information w (t) of the robot in working region is global, i.e., Realize vision positioning.
4. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 3 Method, it is characterized in that, need to shift to an earlier date setting identification line image processing time interval for visual pattern processing, vision collecting device every Setting time interval is performed once, is handled and obtained and output device people course angle information by visual pattern.
5. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 1 Method, it is characterized in that, the specific method of the step (3) is:
3-1) the point set data that laser radar is collected are transformed under rectangular coordinate system;
Clustering processing 3-2) is carried out to the point set under rectangular coordinate system using clustering distance threshold value index;
3-3) point set after cluster is carried out curve fitting using least square method, reference substance characteristic straight line equation is obtained, really The center point coordinate of fixed each reference substance characteristic straight line;
3-4) adjacent feature straight line is matched, by being carried out to the local feature offset between each reference substance characteristic straight line Average computation is distributed, the optimal offset of consecutive frame feature is drawn;
Previous frame data 3-5) are added into a little optimal offset obtained in the previous step, local map is obtained;Repeat above-mentioned Processing procedure;
3-6) by loop iteration, new laser data point set is matched with legacy data, synchronized update positional information and course Angle information.
6. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 5 Method, it is characterized in that, the step 3-2) specific method be:
The distance between point set consecutive points are calculated by order, judged whether within clustering distance threshold value indication range, if Then consecutive points are clustered in the range of;The isolated point beyond each cluster areas is will be independent of to remove.
7. a kind of composition side of the Intelligent Mobile Robot based on Fusion composition as claimed in claim 5 Method, it is characterized in that, the step 3-4) specific method be:
Front and rear adjacent each reference substance characteristic straight line central point distance of two frames is compared with characteristic distance threshold value, if adjacent feature is straight Line center point coordinate is then considered the adjacent same reference substance characteristic straight line of two frames less than characteristic distance threshold value;Use a later frame reference substance The center point coordinate of characteristic straight line subtracts former frame correspondence reference substance feature point coordinates, obtains local feature offset, will be adjacent Local feature offset between each reference substance characteristic straight line of frame carries out distribution average computation, draws the optimal skew of consecutive frame feature Amount.
CN201510289939.9A 2015-05-29 2015-05-29 Intelligent Mobile Robot and its method based on Fusion composition CN104848851B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510289939.9A CN104848851B (en) 2015-05-29 2015-05-29 Intelligent Mobile Robot and its method based on Fusion composition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510289939.9A CN104848851B (en) 2015-05-29 2015-05-29 Intelligent Mobile Robot and its method based on Fusion composition

Publications (2)

Publication Number Publication Date
CN104848851A CN104848851A (en) 2015-08-19
CN104848851B true CN104848851B (en) 2017-08-18

Family

ID=53848677

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510289939.9A CN104848851B (en) 2015-05-29 2015-05-29 Intelligent Mobile Robot and its method based on Fusion composition

Country Status (1)

Country Link
CN (1) CN104848851B (en)

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106595630B (en) * 2015-10-14 2019-09-24 山东鲁能智能技术有限公司 It is a kind of that drawing system and method are built based on laser navigation Intelligent Mobile Robot
CN106584451B (en) * 2015-10-14 2019-12-10 国网智能科技股份有限公司 automatic transformer substation composition robot and method based on visual navigation
CN105865449A (en) * 2016-04-01 2016-08-17 深圳杉川科技有限公司 Laser and vision-based hybrid location method for mobile robot
CN105700532B (en) * 2016-04-19 2018-08-10 长沙理工大学 The Intelligent Mobile Robot navigator fix control method of view-based access control model
CN105953787A (en) * 2016-04-27 2016-09-21 河北德普电器有限公司 Robot navigation map generation system
CN106052691B (en) * 2016-05-17 2019-03-15 武汉大学 Close ring error correction method in the mobile drawing of laser ranging
CN106020188B (en) * 2016-05-17 2018-10-30 杭州申昊科技股份有限公司 A kind of Intelligent Mobile Robot recharging method based on laser navigation
CN106052674B (en) * 2016-05-20 2019-07-26 青岛克路德机器人有限公司 A kind of SLAM method and system of Indoor Robot
CN106123890A (en) * 2016-06-14 2016-11-16 中国科学院合肥物质科学研究院 A kind of robot localization method of Fusion
CN106325269B (en) * 2016-08-05 2019-06-11 上海电力学院 Two-wheeled balance car course corrections device and method based on odometer and Magnetic Sensor
CN106292660B (en) * 2016-08-05 2019-06-11 上海电力学院 Balance car course corrections device and method based on odometer and gray-scale sensor
CN106352870B (en) * 2016-08-26 2019-06-28 深圳微服机器人科技有限公司 A kind of localization method and device of target
CN106227220A (en) * 2016-09-28 2016-12-14 关健生 Independent navigation crusing robot based on Distributed Architecture
CN106325277A (en) * 2016-09-28 2017-01-11 关健生 Distributed framework based autonomous navigation and inspection robot control method
CN106525025B (en) * 2016-10-28 2019-10-11 武汉大学 A kind of Intelligent Mobile Robot route planning air navigation aid
CN107063264A (en) * 2017-04-13 2017-08-18 杭州申昊科技股份有限公司 A kind of robot map creating method suitable for extensive substation
CN107346133A (en) * 2017-07-04 2017-11-14 武汉视览科技有限公司 A kind of the autonomous of indoor mobile robot builds drawing method and device
CN107765145B (en) * 2017-09-26 2019-12-10 国网智能科技股份有限公司 Automatic partial discharge detection device, system and method
CN107677279A (en) * 2017-09-26 2018-02-09 上海思岚科技有限公司 It is a kind of to position the method and system for building figure
CN107907124A (en) * 2017-09-30 2018-04-13 杭州迦智科技有限公司 The localization method known again based on scene, electronic equipment, storage medium, system
CN108181636A (en) * 2018-01-12 2018-06-19 中国矿业大学 Petrochemical factory's crusing robot environmental modeling and map structuring device and method
CN108646265A (en) * 2018-05-11 2018-10-12 北京华捷艾米科技有限公司 Crusing robot and its method for inspecting, cruising inspection system

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101702010A (en) * 2009-10-20 2010-05-05 河北工业大学 Magnetic field measuring device based on visual guiding robot
WO2010130179A1 (en) * 2009-05-15 2010-11-18 北京工业大学 Flexible two-wheel self-balance robot system and motion control method thereof
CN202153615U (en) * 2011-07-12 2012-02-29 重庆大学 Robot for transformer station device inspection tour
CN102789234A (en) * 2012-08-14 2012-11-21 广东科学中心 Robot navigation method and robot navigation system based on color coding identifiers
CN103234542A (en) * 2013-04-12 2013-08-07 东南大学 Combination vehicle curve driving track measurement method base on visual sense
CN103292804A (en) * 2013-05-27 2013-09-11 浙江大学 Monocular natural vision landmark assisted mobile robot positioning method
CN103644904A (en) * 2013-12-17 2014-03-19 上海电机学院 Visual navigation method based on SIFT (scale invariant feature transform) algorithm
CN203719666U (en) * 2013-11-21 2014-07-16 西安中科光电精密工程有限公司 Combined navigation system
CN104470685A (en) * 2013-01-18 2015-03-25 艾罗伯特公司 Mobile Robot Providing Environmental Mapping For Household Environmental Control

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010130179A1 (en) * 2009-05-15 2010-11-18 北京工业大学 Flexible two-wheel self-balance robot system and motion control method thereof
CN101702010A (en) * 2009-10-20 2010-05-05 河北工业大学 Magnetic field measuring device based on visual guiding robot
CN202153615U (en) * 2011-07-12 2012-02-29 重庆大学 Robot for transformer station device inspection tour
CN102789234A (en) * 2012-08-14 2012-11-21 广东科学中心 Robot navigation method and robot navigation system based on color coding identifiers
CN104470685A (en) * 2013-01-18 2015-03-25 艾罗伯特公司 Mobile Robot Providing Environmental Mapping For Household Environmental Control
CN103234542A (en) * 2013-04-12 2013-08-07 东南大学 Combination vehicle curve driving track measurement method base on visual sense
CN103292804A (en) * 2013-05-27 2013-09-11 浙江大学 Monocular natural vision landmark assisted mobile robot positioning method
CN203719666U (en) * 2013-11-21 2014-07-16 西安中科光电精密工程有限公司 Combined navigation system
CN103644904A (en) * 2013-12-17 2014-03-19 上海电机学院 Visual navigation method based on SIFT (scale invariant feature transform) algorithm

Also Published As

Publication number Publication date
CN104848851A (en) 2015-08-19

Similar Documents

Publication Publication Date Title
AU2006203491B2 (en) Method for processing stereo vision data using image density
JP2015181042A (en) detection and tracking of moving objects
US20130135474A1 (en) Automotive Camera System and Its Calibration Method and Calibration Program
CN103941746B (en) Image processing system and method is patrolled and examined without man-machine
CN103868460B (en) Binocular stereo vision method for automatic measurement based on parallax optimized algorithm
KR20070090194A (en) Image recognizing apparatus and method, and position determining apparatus, vehicle controlling apparatus and navigation apparatus using the image recognizing apparatus or method
KR20060013640A (en) Automatic processing of aerial images
EP3008708B1 (en) Vision augmented navigation
CN102426019B (en) Unmanned aerial vehicle scene matching auxiliary navigation method and system
CN103413313B (en) The binocular vision navigation system of electrically-based robot and method
Pink et al. Visual features for vehicle localization and ego-motion estimation
CN103411553B (en) The quick calibrating method of multi-linear structured light vision sensors
CN102789234B (en) Robot navigation method and robot navigation system based on color coding identifiers
CN103400392B (en) Binocular vision navigation system and method based on Intelligent Mobile Robot
CN103236160B (en) Road network traffic condition monitoring system based on video image processing technology
CN105260699B (en) A kind of processing method and processing device of lane line data
JP5588812B2 (en) Image processing apparatus and imaging apparatus using the same
Ozgunalp et al. Multiple lane detection algorithm based on novel dense vanishing point estimation
CN101109640A (en) Unmanned aircraft landing navigation system based on vision
CN104282020B (en) A kind of vehicle speed detection method based on target trajectory
CN102346015A (en) Method for measuring icing thickness of insulator of power transmission line based on video difference analysis
CN104704384A (en) Image processing method, particularly used in a vision-based localization of a device
CN107750364A (en) Detected using the road vertically profiling of stable coordinate system
CN101619984B (en) Mobile robot visual navigation method based on colorful road signs
CN103955920A (en) Binocular vision obstacle detection method based on three-dimensional point cloud segmentation

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
EXSB Decision made by sipo to initiate substantive examination
GR01 Patent grant
GR01 Patent grant
CP03 Change of name, title or address

Address after: 250101 Electric Power Intelligent Robot Production Project 101 in Jinan City, Shandong Province, South of Feiyue Avenue and East of No. 26 Road (ICT Industrial Park)

Patentee after: National Network Intelligent Technology Co., Ltd.

Address before: 250101 Shandong city of Ji'nan province high tech Zone (Lixia District) Xinluo Avenue No. 2008 silver bearing building B block 626

Patentee before: Shandong Luneng Intelligent Technology Co., Ltd.

CP03 Change of name, title or address