CN104699104B - A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus - Google Patents
A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus Download PDFInfo
- Publication number
- CN104699104B CN104699104B CN201510117835.XA CN201510117835A CN104699104B CN 104699104 B CN104699104 B CN 104699104B CN 201510117835 A CN201510117835 A CN 201510117835A CN 104699104 B CN104699104 B CN 104699104B
- Authority
- CN
- China
- Prior art keywords
- agv
- camera
- picture
- angle
- lines
- 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.)
- Expired - Fee Related
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, support frame is anchored on AGV car bodies in described device, steering wheel is fixed on support frame, steering wheel is fixedly connected with runing rest, camera is fixed on runing rest, the camera is located on AGV Vehicular body fronts axis, and camera can adjust sight gear angle by the drive of steering wheel.According to the change in location that line width variations, guiding lines center line and picture bottom and picture middle part intersection point are guided in camera picture, judge posture of the AGV car bodies relative to guiding lines, camera sight gear angle is adjusted in real time with reference to AGV car body actual motion speed, on the premise of ensuring to guide lines tracking accuracy, bigger actual motion velocity interval is adapted to.The present invention can be according to the sight angle of the speed adjust automatically camera of AGV car bodies, and camera view is open, and navigator fix control is accurate, so as to improve operating efficiency.
Description
Technical field
The invention belongs to technical field of logistic equipment, and in particular to a kind of adaptive AGV vision guided navigation sight adjusting apparatus
Stitching tracking.
Background technology
AGV is a kind of circuit by setting or the map of setting, under the control of logistic dispatching system or manual command, from
Dynamic traveling draws body feed tank automatically, completes material transfer and the unmanned material handling vehicle of handling.Visual guidance by
In containing much information, line trace and Navigation Control can be realized well in complex environment, in addition, being also provided simultaneously with construction
Use cost is low, and it is the master of AGV fundamental researches and application and development with safeguarding outstanding advantages of simple to guide the setting of lines
Want one of field.
Existing visual guidance scheme is using the camera observation AGV guide lines being fixedly mounted, and setting angle is mostly
Select perpendicular to the ground, to obtain the video pictures of no perspective effect, to reach the purpose for improving vision positioning precision.For at a high speed
For AGV, in the case of the car body speed of service is higher, due to camera video frame per second be generally fixed for 25 frames/second or 30 frames/
Second, the displacement of inter picture is possible to the field range that can exceed video pictures so that machine vision is entirely ineffective；Therefore,
To ensure the validity of machine vision work, it is necessary to which the speed of service according to AGV car bodies is adjusted in real time to camera sight
It is whole, expand or reduce the size in the video pictures visual field；Correspondingly, the installation form of camera also should not be only perpendicular to ground
Face but before being tilted towards, adjustable angle.
Chinese patent application publication number CN1438138A, publication date August in 2003 27 days, innovation and creation it is entitled " from
The visual guide method and its automatic guide vehicle of dynamic guide car ", this application case discloses a kind of intake surface operating path
Graticule, the processing method of station address code identifier and running status control identifier and electric car design.Its deficiency
Part is that single camera is installed on vehicle body vertically downward, and camera view is limited by body width, can only see
The local cell domain of AGV car bodies covering.
Chinese patent application publication No. CN102997910A, date of publication on March 27th, 2013, innovation and creation it is entitled
" one kind is based on road of ground surface target positioning guidance system and method ", this application case eliminates light using light shield and corresponding lighting apparatus
Source is polluted, and vision sensor vertically downward install by ground proximity.It is disadvantageous in that camera mounting distance is too low, and the visual field is narrow
Narrow to cause information content too low, anti-lines are stained ability than relatively limited.
Chinese patent application publication No. CN103529838A, date of publication on January 22nd, 2014, innovation and creation it is entitled
" the multi-vision visual guide drive device and its collaboration scaling method of automatic guided vehicle ", this application case discloses a kind of multiple shootings
Head detection guiding lines drift angle and the apparatus and method of displacement bias.Each camera is respectively positioned on vehicle body, camera periphery
Arbitrary source illumination is arranged, dig-ins ground guiding lines vertically downward.
Chinese patent grant number CN102346483B, authorized announcement date on November 28th, 2012, innovation and creation it is entitled
" the AGV progress control methods based on passive RFID and auxiliary vision ", mandate cloth on the ad-hoc locations such as stop, turning point
LED or charactron is put to observe on both sides of the road LED or digital screen with display control signal, the CCD camera of AGV vehicle body sides installation
Pictorial symbolization, there is provided the image of control instruction.
Chinese Patent Application No. CN102608998A, data of publication of application on July 25th, 2012, innovation and creation it is entitled
" the visual guidance AGV system and method for embedded system ", this application case are simultaneously vertical with one using a forward sight camera
Downward camera, forward sight camera collection distant view image, vertical camera are used for secondary accurate positioning.Forward sight camera and ground
Face tilts installation at an angle, prediction will turn round, stop, position before deceleration etc.；Vertical camera collection is walked
Routing information, complete stitching tracing task.Wherein the setting height(from bottom) of forward sight camera is transported with ground angle by the maximum of dolly
Scanning frequency degree is relevant, and visual field solstics is 1~1.2 times of maximum operational speed.It is disadvantageous in that, forward sight camera with it is vertical
The setting angle of camera is fixed angle, and fail to provide setting height(from bottom), ground angle and dolly maximum operational speed it
Between physical relationship；Dimension between car body maximum operational speed and visual field solstics distance is different, and chronomere is unclear, it is difficult to
The specific size of visual field far point distance AGV car bodies is specified, it is random big；Critical weak point is also to be, vertical shooting
Head is close to ground, only has the less visual field, is only capable of keeping a close watch on the lines part of very small part, loses a large amount of environmental informations,
It is difficult in adapt to various live local pollutions, such as footprint, greasy dirt, breakage etc..
In summary, there is obvious weakness for existing AGV vision guided navigations scheme：1st, the mount scheme of camera is with vertical
Based on straight installation downwards, camera view is narrow and small, is only limitted to the region of light shield or vehicle body very little, it is difficult to adapts to AGV height
Speed operation；2nd, tilt the camera of installation also in the form of fixed installation based on, lack setting angle and the car body speed of service, navigation
The physical relationship expression formula of positioning control accuracy, field range etc.；Stitching tracking relies primarily on vertical camera and completed；3rd, tilt
The camera of installation is only used for predicting road ahead situation, is not contacted with stitching tracing task directly.
The content of the invention
It is an object of the present invention to provide a kind of stitching track side of adaptive AGV vision guided navigation sight adjusting apparatus
Method, instant invention overcomes the shortcomings that camera view in the prior art is narrow and small, and camera angle can not adjust.
A kind of adaptive AGV vision guided navigation sight adjusting apparatus of the present invention, including：Camera, runing rest, rudder
Machine and support frame, support frame as described above are anchored on AGV car bodies, and steering wheel is fixed on support frame, and steering wheel is fixed with runing rest
Connection, camera are fixed on runing rest, and the camera is located on AGV Vehicular body fronts axis, and camera passes through steering wheel
Drive can adjust sight angle gear.
A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus of the present invention, comprises the following steps：
Picture frame number i=1 is put, receives the instruction of AGV task schedulings and is advanced forward with command speed, wheel turning angle is set
Adjust increment initial value
According to the feedback data of current car body actual motion speed, camera angle gear is adjusted, is calculated under current angular
The auxiliary parameter K of front ground guiding lines,Wherein camera angle is camera and plumb line angle；Wherein：K is represented
A constant related to camera angle gear,Represent axis between video pictures center pixel row and bottom pixel row
Corresponding ground interplanar distance；
Extraction is on two height and positions of the i-th -1 frame video pictures bottom line and middle Central Bank, guiding on lines width
Lines middle separated time and picture bottom and picture are guided in number of pels per line amount between two edges, and the i-th -1 frame video pictures
Location of pixels transverse direction sequence number iBtm and iMid corresponding to two intersection points of central horizontal pixel column, calculate under current angular gear
Attitude Tracking error assessment function J threshold value JεAnd pos is deviateed in the gradient dir of the guide wire in picture and position；
AGV wheel turning angles are arranged to
The currently practical speed of service of AGV car bodies is checked, camera angle gear is adjusted, if angle gear changes, weighs
The new auxiliary parameter K for calculating current angular inferoanterior ground guiding lines,If angle gear is unchanged, turn in next step；
Extract two that lines middle separated time and picture bottom and picture central horizontal pixel column are guided in the i-th frame video pictures
Location of pixels transverse direction sequence number corresponding to intersection point, the gradient dir and position for calculating the guide wire in picture deviate pos, enter one
Step calculates Attitude Tracking error assessment function J and evaluation function increment Delta J；
Judge whether evaluation function increment Delta J is less than or equal to Attitude Tracking error assessment function J threshold value Jε, if so,
Judge whether Attitude Tracking error assessment function J is less than or equal to Attitude Tracking error assessment function J threshold value Jε；If it is not, then
Turn in next step；
Calculate angle adjustment amountVideo frame number i=i+1, AGV wheel turning angles are reset, wherein angle is adjusted
Whole amountReturn to and check the currently practical speed of service v of AGV car bodies, adjust camera angle
Gear β, continue next round cycle detection；In above formula：J(i)It is Attitude Tracking error assessment functional value, Δ corresponding to the i-th frame picture
J(i)It is Attitude Tracking error assessment functional value J corresponding to the i-th frame picture(i)Increment；
Judge whether Attitude Tracking error assessment function J is less than or equal to Attitude Tracking error assessment function J threshold value Jε,
If so, then keeping wheel turning angle constant, video frame number i=i+1, return to and check the currently practical operation speed of AGV car bodies
V is spent, adjusts camera angle gear β, continues next round cycle detection；If it is not, then turn in next step；
Calculate angle adjustment amountThe angle adjustment amountVideo frame number i=i+1,
AGV wheel turning angles are reset,Return and check the currently practical speed of service v of AGV car bodies, adjustment is taken the photograph
As brilliance degree gear β, continue next round cycle detection, in above formula：pos(i)It is position offset corresponding to the i-th frame picture, T is
Frame of video refresh cycle, v are the currently practical speeds of service of AGV car bodies.
Further, angle gear of the sight adjustment according to the selection of AGV current kinetics speed so that two continuous frames regard
Between frequency picture, AGV car bodies displacement is no more than video picturesBreadth.
Further, the pixel count on lines width between two edges in a line will be guided in real-time video picture
Amount, as lateral displacement amount calculating benchmark in the video pictures under different angle gear.
Further, according to the center line and video pictures center pixel row and bottom pixel that lines are guided in video pictures
Location of pixels transverse direction sequence number iMid and iBtm corresponding to two capable intersection points, construction AGV vehicle body attitudes tracking error evaluation letter
Number J=K2·dir2+pos2As adjustment AGV car body steering angle directions and the foundation of size so that AGV postures meet stitching
Tracking accuracy requirement, wherein：Dir=iMid-iBtm,In above formula：K is relevant with camera angle gear β
Constant, dir be picture in guide wire gradient, represent car body and guiding lines between gradient, pos be position skew
Amount, represent car body and offset relative to the position of guiding lines, n is video pictures transverse width resolution ratio.
Further, AGV car bodies turn to adjustment mode and realized by way of adjusting AGV wheel turning angles.
Further, AGV car bodies turn to adjustment mode and realized by the differential mode in AGV wheels both sides.
The beneficial effects of the present invention are：The present invention can be according to the angle of sight of the speed adjust automatically camera of AGV car bodies
Degree, camera view is open, and navigator fix control is accurate, so as to improve operating efficiency.
Brief description of the drawings
Fig. 1 show a kind of adaptive AGV vision guided navigation sight adjusting apparatus structural representations of the present invention.
Fig. 2 show a kind of stitching tracking flow chart of adaptive AGV vision guided navigation sight adjusting apparatus of the present invention.
Fig. 3 show every in a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus of the present invention
Parameter schematic diagram.
Embodiment
The present invention is described in detail below in conjunction with specific embodiment.It should be noted that the skill described in following embodiments
The combination of art feature or technical characteristic is not construed as isolated, and they can be mutually combined so as to reach more preferable
Technique effect.
As shown in figure 1, adaptive AGV vision guided navigation sight adjusting apparatus of the present invention, including：Camera 1, rotation
Support 2, steering wheel 3 and support frame 4, support frame as described above 4 are anchored on AGV car bodies 5, and steering wheel 3 is fixed on support frame 4, steering wheel 3
It is fixedly connected with runing rest 2, camera 1 is fixed on runing rest 2, and the camera 1 is located at 5 anterior axis of AGV car bodies
On line 7, camera 1 can adjust sight angle gear by the drive of steering wheel 3.
The adjustment of the adaptive AGV vision guided navigation sight adjusting apparatus sights is by adjusting the angle shelves of forward sight camera 1
Position is completed, relevant with the current kinetic speed of AGV car bodies 5.When movement velocity is relatively low, less sight angle gear is selected to obtain
Get Geng Gao positional precision；When movement velocity is higher, select larger sight angle gear to obtain bigger visual field breadth, protect
Demonstrate,prove the stitching pursuit movement precision of AGV car bodies 5.
Camera 1 sets some angle gears, and Typical angles gear has 4 gears, respectively 30 °, 45 °, 60 ° and 75 °.
A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus provided by the invention, including it is as follows
Step：
Step S1：Picture frame number i=1 is put, receives the instruction of AGV task schedulings and is advanced forward with command speed, car is set
Take turns steering angle adjustment increment initial value
Step S2：According to the current actual motion speed v of car body 5 feedback data, the angle gear β of camera 1 is adjusted, is calculated
The auxiliary parameter K of current angular inferoanterior ground guiding lines 8,The wherein angle gear β of camera 1 be camera sight with
Plumb line angle, wherein：K represents a constant related to camera angle gear,Represent video pictures center pixel row
The axis interplanar distance correspondingly between bottom pixel row.
In view of β angles it is larger when, surface road will can only occupy down half of picture, and more than horizon meaningless picture accounts for
According to upper half of picture, laterally or longitudinally location of pixels change of the guiding lines 8 in adjacent two frames continuous videos picture is taken, must not
More than video pixel lower half breadthPreferred value isEnsure in continuous two frames video pictures, AGV car bodies 5 relative to
The location variation of lines 8 is guided without the breadth size that may exceed video, can correctly be captured.According to current AGV
The speed of service v of car body 5 size, the angle gear β of camera 1 is adjusted by steering wheel 3 so thatNot less than vT, whereinFor
Surface road length corresponding to video pictures lower half breadth height, T are the time for obtaining 1 frame picture.
The angle gear β of camera 1 adjustment can make it that the compression of guiding lines 8 distorts with other pictures caused by perspective effect
Among dynamic changing process, therefore traditional image calibration technology will fail.Must also can only be completely empty in video pictures
Between in, according to guiding lines 8 itself as with reference to progress AGV stitching tracing controls.
Grey level histogram enhancing is carried out to video pictures, two edges of guiding lines 8 are extracted using Laplace operator,
The straightway starting point that guiding lines or so two edges are represented in video pictures is calculated with pointing to using straight line Hough transformation
Angle, the middle separated time 9 of two edge lines is taken as the vector mark of guiding lines 8.
Ground vertical interval between video pictures center pixel row and bottom pixel row isWherein H is height of the pivot of camera 1 apart from ground, and α is the left and right of camera 1
Visible angle, representative value are 30 degree.In video pictures, each pixel spacing in every a line is approximate it is considered that equal.Lead
Lead bar 8 uses 45~50mm width criteria light tones adhesive tape such as white, yellow, pink, be directly adhered to dark ground and
Into.The number of pixels that is presented in video pictures of developed width of guiding lines 8 only with the angle gear β of camera 1 and shooting
First 1 screen resolution m × n is relevant, can be calculated in advance before AGV operations.
Wherein auxiliary parameterFor a constant relevant with the angle beta gear of camera 1,
Wherein d is distance of the car body front-wheel 6 apart from the front end of car body 5, and L is the distance that car body front-wheel 5 arrives trailing wheel 10.
Step S3：Extraction is on two height and positions of the i-th -1 frame video pictures bottom line and middle Central Bank, guiding lines 8 are wide
Between degree direction Shang Liangge edges lines middle separated time 9 and picture bottom are guided in number of pels per line amount, and the i-th -1 frame video pictures
Location of pixels transverse direction sequence number corresponding to two intersection points of portion and picture central horizontal pixel column, calculate under current angular gear β
The gradient of guiding lines in the threshold value and picture of Attitude Tracking error assessment function and position are deviateed；
Number of pels per line amount is designated as on line segment between two edges on the guiding width of lines 8 in certain frame picture
iPixNum。
Grey level histogram enhancing is carried out to the i-th -1 frame video pictures, corrodes expansion process, takes Hough transformation to extract edge
Straight line, extract on two height and positions of bottom line and middle Central Bank in first width picture, guide two edges on the width of lines 8
Between number of pels per line amountWithAnd i-th -1 guide lines middle separated time in frame video pictures
9 location of pixels transverse direction the sequence number iBtm and iMid corresponding to two intersection points of picture bottom and picture central horizontal pixel column.
According to fixing certain angle gear β, in the bottom line and picture in certain frame picture on two height positions of Central Bank,
The corresponding horizontal actual distance on the ground of number of pels per line amount on the guiding width of lines 8 is calculated, is designated as respectivelyWithThe pixel line width constant being recorded as under respective notch sight angle β.
Guide lines middle separated time 9 and the pixel corresponding to two intersection points of picture bottom and picture central horizontal pixel column
Position transverse direction sequence number, is designated as iBtm and iMid respectively.AGV course angle θ, θ is the anterior axis 7 of AGV car bodies 5 and guiding lines
The angle of middle separated time 9, meet：
During stitching tracking, the two row location of pixels difference expression formulas for taking AGV course angles θ are dir=iMid-iBtm.According to
According to the course angle required precision of θ≤2 °, the two row location of pixels differences corresponding to course angle threshold value areThat is course angle dir≤εθ。
During stitching tracking, the pixel expression formula of AGV position deviations is taken asFoundation ± 20mm's
Position accuracy demand, ε is taken as corresponding to the pixel threshold of AGV positional precisionsD=0.4iPixNum0, i.e. position deviation pos≤
εD。
The threshold value for taking Attitude Tracking error assessment function is less one among both, both Jε=min { εθ,εD}。
The gradient of guiding lines 8 in picture and position, which are deviateed, is respectively：
It is J=K to make AGV Attitude Tracking error assessments function2·dir2+pos2, the posture of AGV car bodies 5 that and if only if just with
When guiding lines 8 are completely superposed, tracking error J is just zero.
Step S4：AGV wheel turning angles are arranged to
To cause Attitude Tracking error assessment function J to reduce about 10% as principle, Newton-decline method is taken to calculate per frame
AGV wheel angle adjustment amounts in picture, the non_derivative changed with the pixel in two continuous frames picture calculate, then and the 3rd
Angle adjustment amount in frame picture isIt is therein to include
Number subscript (i) represents the sequence number of frame of video.Per between the switching of frame picture, wheel turning angle is constantly adjustedUntil tracking error J
Within the scope of designated precision, ensure course error dir≤εθ, and position error pos≤εD, meet that Attitude Tracking error is commented
The threshold value J of valency functionε=min { εθ,εD}。
Step S5：5 currently practical speed of service v of AGV car bodies is checked, camera angle gear β is adjusted, if gear β has change
Change, then go to step S2 recalculate current angular inferoanterior ground guiding lines 8 auxiliary parameter K,If gear β is unchanged,
Then turn in next step.
Step S6：Extract and lines middle separated time 9 and picture bottom and picture central horizontal pixel are guided in the i-th frame video pictures
Location of pixels transverse direction sequence number iBtm corresponding to two capable intersection points(i)And iMid(i), calculate inclining for the guiding lines 8 in picture
Gradient dir(i)Deviate pos with position(i), further calculate attitude error evaluation function J(i)With evaluation function increment Delta J(i)。
The i-th frame video pictures are gathered, grey level histogram enhancing is carried out to the i-th frame picture, corrodes expansion process, takes Hough
Straight line conversion extraction edge line, extracts iBtm(i)、iMid(i)。
The gradient of guiding lines 8 in picture and position, which are deviateed, is
Attitude error evaluation function and evaluation function increment are respectively：
J(i)=K2·dir(i)·dir(i)+pos(i)·pos(i), Δ J(i)=K2·dir(i)·Δdir(i)+pos(i)·Δ
pos(i)。
Step S7：Judge whether evaluation function increment is less than or equal to the threshold value of Attitude Tracking error assessment function, that is, sentence
Disconnected Δ J(i)≤JεWhether set up, if going to step S9, otherwise turn in next step.
Step S8：Calculate angle adjustment amountVideo frame number i=i+1, go to step S4 and reset AGV wheels turn
To angle,
Wherein angle adjustment amount
Step S9：Judge Attitude Tracking error assessment function J(i)Whether Attitude Tracking error assessment function is less than or equal to
Threshold value Jε, that is, judge J(i)≤JεWhether set up, if so, turning in next step；Otherwise S11 is gone to step.
Step S10：Keep wheel turning angle constant, i.e.,Video frame number i=i+1, goes to step S4, due toThat is, going to step S5 reexamines the currently practical speed of service v of AGV car bodies 5, camera angle gear is adjusted
β。
Step S11：Calculate angle adjustment amountVideo frame number i=i+1, go to step S4 and reset AGV wheels
Steering angle.
Wherein angle adjustment amount
It is by constantly adjusting AGV both sides wheel turning angle and constantly adjustment that AGV car bodies 5 of the present invention, which turn to adjustment mode,
What the mode of AGV both sides wheel differential was realized.It is so-called differential, it is exactly that the wheel rotor speed on car body both sides differs, or rotating speed phase
Instead；When both sides vehicle wheel rotational speed difference, car body will turn to the wheel direction of rotating speed slow some or opposite direction rotation.
With reference to experiment, further checking effectiveness of the invention, experimental procedure are as follows：
Ground guiding lines 8 are taken as yellow standard rubber band, width 50mm, are directly adhered to dark ground.AGV car bodies 5
Width is 380mm, and wheelbase 600mm, camera 1 is installed on the front axis 7 of car body 5, positioned at the axial line of AGV front-wheels 6
100mm before, both d=100mm, L=600mm.Camera 1 is directly driven by steering wheel 3, all the time in four fixed angle shelves
Some particular location among position, angle gear β is respectively 30 °, 45 °, 60 ° and 75 °.The setting height(from bottom) of camera 1 is H=
300mm；The video breadth of camera 1 is 640 × 480, and frame per second is 25 frames/second, and left and right angle of visibility size is α=30 °.Corresponding to not
With angle gear, in theory the true breadth in ground corresponding to video pictures should be respectively：
| β | 30° | 45° | 60° | 0 |
| Corresponding ground width among picture | 180mm | 220mm | 311mm | 600mm |
| Picture bottom corresponding ground width | 161mm | 180mm | 220mm | 311mm |
| To bottom corresponding ground height among picture | 93mm | 127mm | 220mm | 600mm |
But also have with short transverse size due to the actual left and right angle of visibility α of camera 1 and inaccurately for 30 ° and width
Difference, therefore, the initial value during gross data is only write as algorithm in upper table set reference data.Real data is according to this hair
When AGV described in bright starts, guiding lines 8 carry out pixel counts approximately perpendicular to the original state of video pictures bottom, write from memory
It is 50mm to recognize line thickness, and the center row pixel in video pictures frame is counted with bottom line number of pixels respectively, is designated as
IPixNum under current angular gear0With
AGV car bodies 5 are placed in when flat on guiding lines 8, guiding lines 8 appear in camera picture near normal to
Before.The instruction of AGV task scheduling systems requires, starts AGV car bodies 5 and runs forward.The adjustment of the angle of camera 1 is according to AGV car bodies 5
Current real-time speed feeds back and carried out.Among ensuring the running of AGV car bodies 5, the interframe displacement for guiding lines 8 is enough
It is small, remove half range face pixels tallBoth 24 pixels, the speed stage threshold value as the angle adjustment of camera 1：
Work as speedWhen, the adjustment of camera angle gear is arranged to 30 °；
Work as speedWhen, the adjustment of camera angle gear is arranged to
45°；
Work as speedWhen, the adjustment of camera angle gear is arranged to
60°；
Work as speedWhen, the adjustment of camera angle gear is arranged to
75°。
Both brought into operation stitching trace routine after the startup of AGV car bodies 5, and grey level histogram enhancing is carried out to video pictures frame,
Corrode expansion process, take Hough transformation to extract edge line, it is straight to the edge on picture middle position and bottom position respectively
Picture between line is counted, and is obtainedWithFurther try to achieve AGV Attitude Tracking error assessment letters
Number threshold valueWith initial angle gear β=30 ° and
Theory of correspondences value is reference, under initial angle gear β=30 °, K=6.18,AndWith
Reference value should be respectivelyWithCorresponding Jε=min { 5100,6336 }=5100.
In follow-up each frame picture processing, grey level histogram enhancing is carried out per frame, corrodes expansion process, takes Hough transformation
Edge line is extracted, is calculatedWithFurther Calculation Estimation function
J(i)=K2·dir(i)·dir(i)+pos(i)·pos(i)With evaluation function increment Delta J(i)=K2·dir(i)Δdir(i)+pos(i)
Δpos(i)；If Δ J(i)More than evaluation function threshold value Jε, then image frame switching between corner adjustment amount beIf Δ J(i)Less than or equal to evaluation function threshold value Jε, then the corner tune between image frame switches
Whole amount isCar body continues to adjust wheel turning angle in the process of runningSize, until evaluate letter
Number J(i)≤Jε, showing that car body 5 is on correct circuit, holding wheel turning angle is constant, and both corner adjustment amount perseverance was zero,
Realize continual and steady stitching tracking.
Although having been presented for some embodiments of the present invention herein, it will be appreciated by those of skill in the art that
Without departing from the spirit of the invention, the embodiments herein can be changed.Above-described embodiment be it is exemplary, no
Restriction that should be using the embodiments herein as interest field of the present invention.
Claims (6)
1. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, the adaptive AGV vision guided navigations regard
Line adjusting device includes：Camera, runing rest, steering wheel and support frame, support frame as described above are anchored on AGV car bodies, steering wheel
It is fixed on support frame, steering wheel is fixedly connected with runing rest, and camera is fixed on runing rest, and the camera is located at
On AGV Vehicular body fronts axis, camera can adjust sight angle gear by the drive of steering wheel；It is characterized in that：It is described from
The stitching tracking of AGV vision guided navigation sight adjusting apparatus is adapted to, is comprised the following steps：
Picture frame number i=1 is put, receives the instruction of AGV task schedulings and is advanced forward with command speed, set wheel turning angle to adjust
Increment initial value
According to the feedback data of current car body actual motion speed, camera angle gear is adjusted, calculates current angular inferoanterior
The auxiliary parameter K of ground guiding lines,Wherein camera angle is camera and plumb line angle；Wherein：K refers to image
A related constant of brilliance degree gear,Refer to that axis is corresponding between video pictures center pixel row and bottom pixel row
Ground interplanar distance；
Extraction is on two height and positions of the i-th -1 frame video pictures bottom line and middle Central Bank, guiding on lines width two
Lines middle separated time and picture bottom and picture center are guided in number of pels per line amount between edge, and the i-th -1 frame video pictures
Location of pixels transverse direction sequence number iBtm and iMid corresponding to two intersection points of horizontal lines, calculate posture under current angular gear
Tracking error evaluation function J threshold value JεAnd pos is deviateed in the gradient dir of the guide wire in picture and position；
AGV wheel turning angles are arranged to
The currently practical speed of service of AGV car bodies is checked, camera angle gear is adjusted, if angle gear changes, counts again
The auxiliary parameter K of calculation current angular inferoanterior ground guiding lines,If angle gear is unchanged, turn in next step；
Extract and lines middle separated time and picture bottom and two intersection points of picture central horizontal pixel column are guided in the i-th frame video pictures
Corresponding location of pixels transverse direction sequence number, calculates the gradient dir of the guide wire in picture and pos is deviateed in position, further meter
Calculate Attitude Tracking error assessment function J and evaluation function increment Delta J；
Judge whether evaluation function increment Delta J is less than or equal to Attitude Tracking error assessment function J threshold value Jε, if so, judging appearance
Whether state tracking error evaluation function J is less than or equal to Attitude Tracking error assessment function J threshold value Jε；If it is not, then calculate angle
Spend adjustment amountVideo frame number i=i+1 is set, resets AGV wheel turning angles, wherein angle adjustment amountReturn to and check the currently practical speed of service v of AGV car bodies, adjust camera angle gear
β, continue next round cycle detection；In above formula：J(i)It is Attitude Tracking error assessment functional value, Δ J corresponding to the i-th frame picture(i)
It is Attitude Tracking error assessment functional value J corresponding to the i-th frame picture(i)Increment；
Judge whether Attitude Tracking error assessment function J is less than or equal to Attitude Tracking error assessment function J threshold value Jε, if so,
Then keep wheel turning angle constant, video frame number i=i+1, return to and check the currently practical speed of service v of AGV car bodies, adjust
Whole camera angle gear β, continue next round cycle detection；If it is not, then calculate angle adjustment amountThe angle adjustment
AmountVideo frame number i=i+1 is set, resets AGV wheel turning angles,Return and check the currently practical speed of service v of AGV car bodies, adjust camera angle gear β, continue next
Cycle detection is taken turns, in above formula：pos(i)It is position offset corresponding to the i-th frame picture, T is the frame of video refresh cycle, and v is AGV cars
The currently practical speed of service of body.
2. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, its feature exist as claimed in claim 1
In：Angle gear of the sight adjustment according to the selection of AGV current kinetics speed so that between two continuous frames video pictures, AGV
Car body displacement is no more than video picturesBreadth.
3. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, its feature exist as claimed in claim 1
In：The pixel quantity on lines width between two edges in a line will be guided in real-time video picture, as different angles
Spend lateral displacement amount calculating benchmark in the video pictures under gear.
4. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, its feature exist as claimed in claim 1
In：Two intersection point institutes according to the center line and video pictures center pixel row and bottom pixel row that lines are guided in video pictures
Corresponding location of pixels transverse direction sequence number iMid and iBtm, construction AGV vehicle body attitude tracking error evaluation functions J=K2·dir2+
pos2As adjustment AGV car body steering angle directions and the foundation of size so that AGV postures meet the requirement of stitching tracking accuracy,
Wherein：Dir=iMid-iBtm,In above formula：K is the constant relevant with camera angle gear β, and dir is
The gradient of guide wire in picture, gradient between car body and guiding lines is represented, pos is position offset, represents car
Body phase is offset for guiding the position of lines, and n is video pictures transverse width resolution ratio.
5. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, its feature exist as claimed in claim 1
In：AGV car bodies turn to adjustment mode and realized by way of adjusting AGV wheel turning angles.
6. a kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus, its feature exist as claimed in claim 2
In：AGV car bodies turn to adjustment mode and realized by the differential mode in AGV wheels both sides.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510117835.XA CN104699104B (en) | 2015-03-17 | 2015-03-17 | A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510117835.XA CN104699104B (en) | 2015-03-17 | 2015-03-17 | A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104699104A CN104699104A (en) | 2015-06-10 |
| CN104699104B true CN104699104B (en) | 2018-02-02 |
Family
ID=53346332
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510117835.XA Expired - Fee Related CN104699104B (en) | 2015-03-17 | 2015-03-17 | A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104699104B (en) |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017092008A1 (en) | 2015-12-03 | 2017-06-08 | 华为技术有限公司 | Navigation method and navigation device |
| CN105550676A (en) * | 2016-02-03 | 2016-05-04 | 天津市乐图软件科技有限公司 | Logistics delivery vehicle trace tracking method based on linearity CCD |
| CN105774861B (en) * | 2016-05-09 | 2018-09-18 | 哈工大机器人集团（广州）知识产权投资控股有限公司哈尔滨分公司 | A kind of family's transfer robot |
| US10175658B2 (en) * | 2016-06-29 | 2019-01-08 | GM Global Technology Operations LLC | Systems and methods for sensor platform |
| CN106969766A (en) * | 2017-03-21 | 2017-07-21 | 北京品创智能科技有限公司 | A kind of indoor autonomous navigation method based on monocular vision and Quick Response Code road sign |
| CN107390683B (en) * | 2017-07-14 | 2021-03-02 | 长沙中联消防机械有限公司 | Automatic rail aligning system and method for rail dual-purpose vehicle and fire fighting truck |
| CN109426251B (en) * | 2017-08-29 | 2020-11-06 | 杭州海康机器人技术有限公司 | Running control method and device for robot |
| CN107632602A (en) * | 2017-09-01 | 2018-01-26 | 上海斐讯数据通信技术有限公司 | AGV trolley travelling tracks method for correcting error and system, terrestrial reference Quick Response Code acquisition device |
| CN108153302B (en) * | 2017-12-08 | 2021-04-20 | 深圳市易成自动驾驶技术有限公司 | Guiding method and device of automatic guided vehicle and computer readable storage medium |
| CN108458709B (en) * | 2018-02-22 | 2020-03-24 | 北京航空航天大学 | Airborne distributed POS data fusion method and device based on vision-aided measurement |
| CN109032125B (en) * | 2018-05-31 | 2021-09-10 | 上海工程技术大学 | Navigation method of visual AGV |
| CN109782755B (en) * | 2018-12-27 | 2022-05-31 | 广东飞库科技有限公司 | Method for controlling AGV to calibrate and AGV to calibrate position, computer storage medium and AGV |
| CN111374600B (en) * | 2018-12-28 | 2021-08-24 | 珠海市一微半导体有限公司 | Self-tracing control method of sweeping robot and intelligent sweeping robot |
| CN109857103B (en) * | 2019-01-22 | 2023-04-07 | 秦皇岛天业通联重工科技有限公司 | Control method, device and system for automatically driving vehicle |
| CN109947137B (en) * | 2019-03-19 | 2021-10-22 | 北京特种机械研究所 | Automatic following device of PGV visual scanning head |
| CN110147096A (en) * | 2019-04-01 | 2019-08-20 | 江苏大学 | A kind of multimachine control method for coordinating that view-based access control model follows |
| CN110209162A (en) * | 2019-05-06 | 2019-09-06 | 珠海达明科技有限公司 | A kind of 4 wheel driven AGV autonomous tracing in intelligent vehicle based on Ackermann steering principle |
| CN110147100A (en) * | 2019-05-06 | 2019-08-20 | 珠海达明科技有限公司 | A kind of AGV platform and air navigation aid with high-precision navigation locating function |
| CN110030932B (en) * | 2019-05-24 | 2020-12-15 | 广东嘉腾机器人自动化有限公司 | AGV deviation measurement method and AGV deviation measurement device |
| CN111674800B (en) | 2020-06-03 | 2021-07-09 | 灵动科技（北京）有限公司 | Intelligent warehousing technology for automatic driving system |
| WO2023077255A1 (en) * | 2021-11-02 | 2023-05-11 | 深圳市大疆创新科技有限公司 | Method and apparatus for controlling movable platform, and movable platform and storage medium |
| CN116437016B (en) * | 2023-06-13 | 2023-10-10 | 武汉中观自动化科技有限公司 | Object scanning method, device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11175147A (en) * | 1997-12-09 | 1999-07-02 | Shinko Electric Co Ltd | Positioning device for moving body |
| CN101183265A (en) * | 2007-11-15 | 2008-05-21 | 浙江大学 | Automatic guidance system based on radio frequency identification tag and vision and method thereof |
| CN102608998A (en) * | 2011-12-23 | 2012-07-25 | 南京航空航天大学 | Vision guiding AGV (Automatic Guided Vehicle) system and method of embedded system |
| CN104085313A (en) * | 2014-07-18 | 2014-10-08 | 安徽中家智锐科技有限公司 | Eight-freedom-degree mechanical arm system of AGV chassis |
| CN204480047U (en) * | 2015-03-17 | 2015-07-15 | 武汉纺织大学 | A kind of self-adaptation AGV vision guided navigation sight line adjusting gear |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060064212A1 (en) * | 2004-09-22 | 2006-03-23 | Cycle Time Corporation | Reactive automated guided vehicle vision guidance system |
-
2015
- 2015-03-17 CN CN201510117835.XA patent/CN104699104B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11175147A (en) * | 1997-12-09 | 1999-07-02 | Shinko Electric Co Ltd | Positioning device for moving body |
| CN101183265A (en) * | 2007-11-15 | 2008-05-21 | 浙江大学 | Automatic guidance system based on radio frequency identification tag and vision and method thereof |
| CN102608998A (en) * | 2011-12-23 | 2012-07-25 | 南京航空航天大学 | Vision guiding AGV (Automatic Guided Vehicle) system and method of embedded system |
| CN104085313A (en) * | 2014-07-18 | 2014-10-08 | 安徽中家智锐科技有限公司 | Eight-freedom-degree mechanical arm system of AGV chassis |
| CN204480047U (en) * | 2015-03-17 | 2015-07-15 | 武汉纺织大学 | A kind of self-adaptation AGV vision guided navigation sight line adjusting gear |
Non-Patent Citations (4)
| Title |
|---|
| AGV视觉导航系统的研究;马道明;《中国优秀硕士学位论文全文数据库信息科技辑（月刊）》;20120215(第02期);I140-264 * |
| 单轮驱动/转向AGV轨迹分析及视觉导航;赵玉麟;《中国优秀硕士学位论文全文数据库信息科技辑（月刊）》;20070715(第01期);I140-49 * |
| 搬运机器人AGV的智能导航研究;金超;《中国优秀硕士学位论文全文数据库信息科技辑（月刊）》;20131215(第S2期);I140-334 * |
| 自动导引车的视觉导航;任小芳;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑（月刊）》;20050715(第03期);I140-47 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104699104A (en) | 2015-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104699104B (en) | A kind of stitching tracking of adaptive AGV vision guided navigation sight adjusting apparatus | |
| CN105511462B (en) | A kind of AGV air navigation aids of view-based access control model | |
| CN105955259B (en) | The accurate positioning method of monocular vision AGV based on multiwindow real time distance | |
| CN104299240B (en) | Camera calibration method and system for lane shift early warning | |
| CN205024577U (en) | Self -walking -type building machine | |
| CN105404842A (en) | AGV positioning, orientation and speed measurement method based on landmark two-dimensional code | |
| CN110033480B (en) | Aerial photography measurement-based airborne photoelectric system target motion vector estimation method | |
| CN106529587B (en) | Vision course recognition methods based on object detection | |
| CN106969766A (en) | A kind of indoor autonomous navigation method based on monocular vision and Quick Response Code road sign | |
| CN109583409A (en) | A kind of intelligent vehicle localization method and system towards cognitive map | |
| CN103733077A (en) | Device for measuring speed and position of a vehicle moving along a guidance track, method and computer program product corresponding thereto | |
| CN110347160A (en) | A kind of automatic guide vehicle and its air navigation aid based on dual camera barcode scanning | |
| CN106651953A (en) | Vehicle position and gesture estimation method based on traffic sign | |
| CN103729837A (en) | Rapid calibration method of single road condition video camera | |
| WO2020083103A1 (en) | Vehicle positioning method based on deep neural network image recognition | |
| CN107392951A (en) | Remote sensing images high accuracy rapid registering method | |
| CN101774432A (en) | Human control intersection docking operation method | |
| CN105869150A (en) | Movable platform calibration device and calibration method based on visual recognition | |
| CN107831776A (en) | Unmanned plane based on nine axle inertial sensors independently makes a return voyage method | |
| CN101813465A (en) | Monocular vision measuring method of non-contact precision measuring corner | |
| CN106525056A (en) | Method for lane line detection by gyro sensor | |
| CN109239752A (en) | Vehicle positioning system | |
| CN109146958A (en) | A kind of traffic sign method for measuring spatial location based on two dimensional image | |
| CN103400362B (en) | Accident close-range figure and Aerial Images merge the method obtaining clear scene graph mutually | |
| CN110023988B (en) | Method and system for generating a combined overhead image of a roadway |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180202 Termination date: 20210317 |