CN106601076B - A kind of automobile self training device and method based on inertial navigation and area array cameras - Google Patents
A kind of automobile self training device and method based on inertial navigation and area array cameras Download PDFInfo
- Publication number
- CN106601076B CN106601076B CN201611040667.XA CN201611040667A CN106601076B CN 106601076 B CN106601076 B CN 106601076B CN 201611040667 A CN201611040667 A CN 201611040667A CN 106601076 B CN106601076 B CN 106601076B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- formula
- yaw angle
- image
- library
- 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
Classifications
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B19/00—Teaching not covered by other main groups of this subclass
- G09B19/16—Control of vehicles or other craft
- G09B19/167—Control of land vehicles
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Closed-Circuit Television Systems (AREA)
- Image Analysis (AREA)
- Navigation (AREA)
Abstract
The invention discloses a kind of automobile self training device and method based on inertial navigation and area array cameras, including 2 area array cameras, interior client, liquid crystal display and loudspeaker.Area array cameras is used to acquire the image information of vehicle initial position, and system obtains the initial position and initial attitude angle of vehicle by the image information that area array cameras acquires.2.4G radio-frequency module, MCU module and digital strap-down navigation device are integrated in interior client, for obtaining physical location in vehicular motion and attitude angle and the deviation for calculating current vehicle position and ideal position.Liquid crystal display is used to show the offset of vehicle.Loudspeaker is for prompting student.The present invention can make student carry out self-service study, improve learning interest, increase learning efficiency, shorten learning time.
Description
Technical field
The present invention relates to technical field of vehicle detection, and in particular to a kind of automobile based on inertial navigation and area array cameras from
Help training device and method.
Background technique
While saturation with China market for automobiles region, the competition between driving school is also growing more intense.How to reduce cost, mention
The training effectiveness of high student annoyings the operators of each driving school.It, will if can be realized the partial automation of driving school's training
A large amount of manpower can be saved, and is conducive to the further operation and development of driving school.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of automobile self training side based on strap-down navigation and area array cameras
Method drives the study of Self-help vehicle formula library or side parking for student, comprising the following steps:
Step 1,2 area array cameras are separately mounted to the two sides of vehicle headstock, when student starts practice library or side
When parking, the relative positional relationship between vehicle and library or side parking start line is detected to obtain by 2 area array cameras
The initial position of vehicle and initial yaw angle；
Step 2, the coordinate at each turning point that vehicle needs to turn to during library or lateral parking, root are determined
According to obtained vehicle initial position and initial yaw angle, carried out curve fitting using interpolation method to ideal planning driving path；
Step 3, real time position and yaw angle of the vehicle during library or lateral parking are obtained；
Step 4, fallen according to vehicle vehicle centroid position coordinates (X (w), Y (w)) during library or side coil and vehicle body it is horizontal
Pivot angle θ, acquires the ideal path normal online to vehicle body yaw angle theta, which is vehicle location offset d；
Step 5, display vehicle location offset d is carried out by being mounted on the display screen before driver's cabin, if student adjusts vehicle
It is in the right direction, then show screen display vehicle location offset d reduce；If student adjusts the anisotropy of vehicle,
Show screen display vehicle location offset d increase, with reach student drive Self-help vehicle formula fall library or side parking；
Vehicle cab is also equipped with loudspeaker:
If the driving path of vehicle deviates ideal path to the left, loudspeaker issues voice " vehicle deviates to the left "；
If the driving path of vehicle deviates to the right ideal path, loudspeaker issues voice " vehicle deviates to the right ".
Further, the initial position of vehicle is obtained described in step 1 and initial yaw angle includes:
Step 11, by 2 area array cameras with shooting garages reticle image, using the ground reticle image as original image,
And noise reduction pretreatment is carried out to original image, obtain pretreatment image；
Step 12, the color space of pretreatment image is converted into HSV by BGR, and binaryzation is carried out to pretreatment image
Processing, obtains binary image；
Step 13, after to binary image progress edge extracting, the straight line after detection edge extracting in bianry image, and mention
The pixel at straight line is taken out, the coordinate of the pixel is fitted, graticule is in image coordinate system with obtaining garage
Slope k1,k2With intercept d1,d2；
Step 14, the initial position by (formula 1), (formula 2) and (formula 3) calculating vehicle and initial yaw angle, i.e. vehicle are horizontal
Pivot angle θ0With the horizontal distance D of vehicle center point and sideline1,D2:
In formula, aij、bij、cij, m, n, q be calibration coefficient.
Further, interpolation method described in step 2 is specially cubic spline interpolation.
Further, real time position and cross of acquisition vehicle described in step 3 during library or lateral parking
Pivot angle includes:
Step 31, the real-time angular velocity signal ω of vehicle is obtained, and ω is filtered；
Step 32, the real-time yaw angle theta of vehicle is calculated by (formula 4):
θt=θt-1+∫ωdt+θ0(formula 4)
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., ω are the real-time angle of vehicle
Speed signal；
Step 33, the real-time acceleration signal a of vehicle is acquired(B), and to a(B)It is filtered；
Step 34, coordinate X of the vehicle centroid under earth coordinates is calculated by (formula 5) and (formula 6)(w)And Y(w):
X(w)=∫ ∫ (a(B)cosθdt)+X0(formula 5)
Y(w)=∫ ∫ (a(B)sinθdt)+Y0(formula 6)
Wherein, X(w)And Y(w)Respectively coordinate of the vehicle centroid under earth coordinates, θ are the real-time yaw angle of vehicle,
X0、Y0The respectively initial coordinate of vehicle centroid.
Further, the ω described in step 31 is filtered by (formula 7):
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., K are filter factor,
For t moment angular velocity signal estimated value.
The present invention also provides a kind of automobile self training device based on strap-down navigation and area array cameras, drives for student
It sails the study of Self-help vehicle formula and falls library or side parking, including image capture module, default ideal path module, obtain mould in real time
Block obtains offset module, self-service driving module；
Described image acquisition module includes 2 area array cameras, and 2 area array cameras are separately mounted to vehicle by near-car head
Two sides, for when student start practice fall library or side stop when, detect vehicles by 2 area array cameras and pass through 2 faces
Array camera detection vehicle and library or side parking start line between relative positional relationship with obtain vehicle initial position and
Initial yaw angle；
The default ideal path module is every for determining that vehicle needs to turn to during library or lateral parking
Coordinate at a turning point, according to obtained vehicle initial position and initial yaw angle, using interpolation method to ideal planning driving path
It carries out curve fitting；
The real-time acquisition module is for obtaining real time position and sideway of the vehicle during library or lateral parking
Angle；
Vehicle centroid position coordinates (the X for obtaining offset module and being used to be fallen according to vehicle during library or side coil
(w), Y (w)) and vehicle body yaw angle theta, the ideal path normal online to vehicle body yaw angle theta is acquired, which is
Vehicle location offset d；
The self-service driving module includes display screen and loudspeaker, is shown by being mounted on the display screen before driver's cabin
Vehicle location offset d shows that the vehicle location offset d of screen display reduces if student adjusts the in the right direction of vehicle；
If student adjusts the anisotropy of vehicle, show that the vehicle location offset d of screen display increases, drives vehicle to reach student
It is self-service fall library or side parking；
Vehicle cab is also equipped with loudspeaker:
If the driving path of vehicle deviates ideal path to the left, loudspeaker issues voice " vehicle deviates to the left "；
If the driving path of vehicle deviates to the right ideal path, loudspeaker issues voice " vehicle deviates to the right ".
Further, the initial position of vehicle is obtained described in image capture module and initial yaw angle includes:
Step 11, by 2 area array cameras with shooting garages reticle image, using the ground reticle image as original image,
And noise reduction pretreatment is carried out to original image, obtain pretreatment image；
Step 12, the color space of pretreatment image is converted into HSV by BGR, and binaryzation is carried out to pretreatment image
Processing, obtains binary image；
Step 13, straight in image after detection edge extracting after carrying out edge extracting (Caany operator) to binary image
Line, and the pixel at straight line is extracted, it is fitted according to the coordinate of the pixel, graticule is sat in image with obtaining garage
Slope k in mark system1,k2With intercept d1,d2；
Step 14, the initial position by (formula 1), (formula 2) and (formula 3) calculating vehicle and initial yaw angle, i.e. vehicle are horizontal
Pivot angle θ0With the horizontal distance D of vehicle center point and sideline1,D2:
In formula, aij、bij、cij, m, n, q be calibration coefficient.
Further, presetting interpolation method described in ideal path module is specially cubic spline interpolation.
Further, real-time position of acquisition vehicle described in module during library or lateral parking is obtained in real time
It sets and includes: with yaw angle
Step 31, the real-time angular velocity signal ω of vehicle is obtained, and ω is filtered；
Step 32, the real-time yaw angle theta of vehicle is calculated by (formula 4):
θt=θt-1+∫ωdt+θ0(formula 4)
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., ω are the real-time angle of vehicle
Speed signal；
Step 33, the real-time acceleration signal a of vehicle is acquired(B), and to a(B)It is filtered；
Step 34, coordinate X of the vehicle centroid under earth coordinates is calculated by (formula 5) and (formula 6)(w)And Y(w):
X(w)=∫ ∫ (a(B)cosθdt)+X0(formula 5)
Y(w)=∫ ∫ (a(B)sinθdt)+Y0(formula 6)
Wherein, X(w)And Y(w)Respectively coordinate of the vehicle centroid under earth coordinates, θ are the real-time yaw angle of vehicle,
X0、Y0The respectively initial coordinate of vehicle centroid.
Further, the ω described in step 31 is filtered by (formula 7):
Wherein, t is value at the time of vehicle falls in library or side docking process, and t=1,2 ..., K are filter factor,For t
Moment angular velocity signal estimated value.
Compared with prior art, the present invention has following technical effect that
The present invention detects vehicle and library or side parking by being installed on 2 area array cameras of the vehicle by near-car head two sides
The starting relative positional relationship of starting point, and ideal vehicle travel path is generated, strapdown is then utilized in vehicular motion
Navigation device obtains the posture and location information of vehicle in real time, and compares with ideal path, calculates current vehicle position
It sets and the deviation of ideal position, the departure is fed back into student finally by human-computer interaction interface, its is and guided to carry out vehicle body
The amendment of posture and position.The training device and its algorithm proposed through the invention, student are able to carry out self-service study,
Learning interest is improved, learning efficiency is increased, shortens learning time.In view of the huge automobile in China investigates market, energy of the present invention
Enough generate huge economic and social benefit.
Detailed description of the invention
Fig. 1 is hardware connection diagram of the invention；
Fig. 2 is the automobile self training device and its algorithm area array cameras of the invention based on strap-down navigation and area array cameras
Schematic view of the mounting position；
Fig. 3 is vehicle initial position and initial yaw angle algorithm flow chart；
Fig. 4 is image binaryzation effect picture；
Fig. 5 is fitting a straight line display figure；；
Signal before Fig. 6 vehicle route node regulation；
Signal after Fig. 7 vehicle route node regulation；
Fig. 8 body modeling schematic diagram；
Fig. 9 vehicle shift amount calculates schematic diagram；
Figure 10 is the comparison diagram for testing path and ideal path.
Specific embodiment
Embodiment 1
A kind of automobile self training device control algolithm based on strap-down navigation and area array cameras is present embodiments provided, is had
Body step are as follows:
1,2 area array cameras are separately mounted to the two sides that vehicle leans on near-car head, when student starts practice library or side
When parking, the relative positional relationship between vehicle and library or side parking start line is detected to obtain by 2 area array cameras
The initial position of vehicle and initial yaw angle；；
The present invention obtains initial position and initial yaw angle of the vehicle centroid in earth coordinates using machine vision,
The flow chart of algorithm is as shown in figure 3, algorithm involved in the part includes following several big steps:
(1) picture signal is obtained, and carries out image preprocessing
The present invention shoots field graticule with falling library using the camera for being installed on vehicular sideview.The original graph in MCU acquisition place
As after, image is pre-processed first, to reduce the noise of image, and improves the resolution ratio of image.
(2) convert simultaneously binaryzation to the color space of image
The present invention uses the method for convert then binaryzation for the color space of picture.In this example, MCU will scheme
The color space of picture is converted to HSV by BGR, then carries out binaryzation according to given threshold value to the saturation degree channel of image.Two
Effect picture after value is as shown in Figure 4.
(3) place of the falling library graticule in image is fitted, finds out parameter of the garage graticule in image coordinate system
In order to obtain posture and position of the vehicle in earth coordinates, need to obtain garage graticule first in image coordinate system
In parameter, including slope, intercept.The method that the present invention uses are as follows: edge extracting is carried out to binary image, is then examined
Straight line in altimetric image extracts the neighbouring pixel that goes beyond the scope later, then carries out place mark according to the coordinate of these pixels
The center line of line is fitted, and obtains relevant straight line parameter.
In this example, edge detection uses Canny operator, and straight-line detection then mainly utilizes Hough transformation, the ginseng of straight line
Number fitting uses least square method.In this example, the display of the straight line after fitting in the picture is as shown in Figure 5.Pass through tune
The setting angle of whole video camera and position can enable camera take the right angle electrical at site boundary right angle in same piece image,
And two mutually perpendicular sidelines of right angle, and the slope k of this two lines is calculated1,k2With intercept d1,d2。
(4) calculating of the conversion of coordinate system and vehicle initial yaw angle and spatial position
In the previous step, the straight line parameter that MCU is calculated is the parameter in image coordinate system, needs to carry out it
Conversion can just obtain yaw angle theta of the vehicle in earth coordinates0With the horizontal distance D of vehicle center point and sideline1,D2.This hair
The bright yaw angle theta for being calculated by the following formula vehicle0
In formula, aij、bij、cij, m, n, q be calibration coefficient.By acquiring experimental data, these coefficients can be determined
Numerical value.
After obtaining vehicle body yaw angle and the distance in centroid distance two vertical sidelines, that is, it can determine the mass center of vehicle in the earth
The initial yaw angle of coordinate and vehicle in coordinate system provides just for the generation of next step ideal path and the integral calculation in path
Beginning condition.
2, the coordinate at each turning point that vehicle needs to turn to during library or lateral parking is determined, according to
The vehicle initial position arrived and initial yaw angle carry out curve fitting to ideal planning driving path using interpolation method；
In the present embodiment using vehicle fall library or lateral parking section start coordinate as origin, vehicle is estimated with the origin
Needing to carry out 7 steerings in library or while changing trains or buses could be parked in vehicle in garage, fall library or lateral parking field in national standard
Ground acquires this 7 turning points to get the coordinate arrived at each turning point.
The centroid position of vehicle when by considering to start reversing, the present invention utilize a kind of algorithm based on cubic spline interpolation
Generate the reversing path of vehicle.Its specific algorithm is as follows:
The present embodiment artificially marks vehicle by the coordinate position at each steering node by training, then according to
According to the initial yaw angle of vehicle and the initial coordinate position of vehicle centroid, these coordinate points are rotated and deviated, is next
The Cubic Spline Fitting of step diametal curve provides node.Signal before and after vehicle route node regulation is as shown in Figures 6 and 7.
Steering node coordinate in the present embodiment by coach's calibration is following (unit: m):
(0,-0.2)(0.6,0.8)(1.6,2.8)(3,3.8)(3.4,3.9)(3.8,3,98)(4.2,4)
Graphics proportion 1cm=2m is taken, then experimental result is as shown in Figure 10.
3, real time position and yaw angle of the vehicle during library or lateral parking are obtained；
During the practice of subject two, the travel speed of vehicle is generally lower, and the topography of selected exercise floor compared with
It is flat.Therefore, the movement of vehicle during two practice of subject can be considered as uniform motion.Therefore two axis inertial navigations are used, it builds
Vertical Car body model is as shown in Figure 8.Wherein, G point is vehicle centroid position, x(B), y(B)Respectively the horizontal axis of vehicle body coordinate system and
The longitudinal axis, x(w), y(w)Respectively earth coordinates are horizontally and vertically.Here vehicle is obtained using following algorithm to fall during library
Position and yaw angle:
1) obtaining the angular velocity signal ω that digital gyroscope measures influences measurement accuracy since there are zero bias for gyroscope,
Temperature correction is carried out firstly the need of to it.Here temperature correction is realized using the temperature-compensation method based on BP neural network.
Since this class algorithm is not the content that the present invention is stressed, which is not described herein again.Then the digital signal value of acquisition is converted to
Circular measure；
2) original angular velocity signal ω includes noise, needs to be filtered it.Since vehicle is in the even of low speed
Fast driving status, therefore it is believed that useful signal is concentrated mainly on low-frequency range.Here believed using the method angular velocity of differential filtering
Number ω is filtered, and specific formula is as follows:
Wherein, K is filter factor,For t moment angular velocity signal estimated value.By suitably choosing the big of filter factor K
It is small, the Real-Time Filtering of angular velocity signal can be completed.
3) by integrating to the angular velocity signal of acquisition, the yaw angle theta of vehicle is solved, calculation formula is as follows:
θt=θt-1+∫ωdt+θ0, t=1,2 ... (7)
Wherein, θtFor vehicle t moment yaw angle.
The value of yaw angle theta according to above-mentioned formula you can get it vehicle, has thereby determined that the yaw angle of vehicle.
4) the acceleration signal a that accelerometer measures is obtained(B), while it is filtered using differential filtering method,
To remove influence of the noise jamming to measurement accuracy.Specific formula is as follows:
Wherein, K is filter factor,For t moment acceleration signal estimated value.By suitably choosing filter factor K's
The Real-Time Filtering to acceleration signal can be completed in size.
5) the acceleration signal a of above-mentioned acquisition(B)For the acceleration signal under vehicle body coordinate system, in order to ask the position of vehicle
Information, it is necessary first to the acceleration signal under vehicle body coordinate system be converted to the acceleration signal under earth coordinates, that is, ask and add
Speed signal a(B)Projection in earth coordinates, specific formula is as follows:
Wherein,Acceleration respectively in earth coordinates lower edge horizontally and vertically.
By carrying out quadratic integral to above-mentioned obtained acceleration, vehicle centroid that you can get it is under earth coordinates
Coordinate X(w)And Y(w)。X(w)And Y(w)Solution formula it is as follows:
X(w)=∫ ∫ (a(B)cosθdt)+X0 (11)
Y(w)=∫ ∫ (a(B)sinθdt)+Y0 (12)
Wherein, X(w)And Y(w)The ordinate and abscissa that are vehicle centroid under earth coordinates, X0And Y0For vehicle centroid
Initial ordinate and abscissa under earth coordinates.
Using above step and algorithm you can get it vehicle centroid in the position (X for the during of falling library(w),Y(w)) and vehicle cross
Pivot angle θ provides condition for the subsequent amount for calculating vehicle shift ideal path.
4, vehicle falls the calculating of offset during library
Vehicle actual travel path and ideal path of vehicle during the offset for the during of falling library refers to library
Deviation.The centroid position coordinate i.e. (X fallen by the available vehicle of digital strap-down navigation during library(w),Y(w)) and vehicle body
Yaw angle theta makees the online normal of yaw angle by ideal path line, then the length of normal degree is vehicle location offset d,
Its schematic diagram is as shown in Figure 9.Algorithm steps and calculation formula are as follows:
(1) the straight line expression formula L of normal is found out；
The present invention seeks straight line expression formula using known some coordinates and slope, and the expression formula of straight line L is as follows:
(2) the intersection point Q of straight line L and ideal path line S (x) are sought；
Simultaneous straight line S (x) and L, solution obtain intersection point Q coordinate (x, y)
(3) ask Q point to the distance d of G point；
5, offset will by way of progress bar display on liquid crystal display, while progress bar can according to the adjustment of student into
Degree is changed (elongation and shortening), and when the adjustment direction of student is correct, with the increase of adjustment amount, progress bar can shorten；
When the adjustment direction mistake of student, with the increase of adjustment amount, progress bar can extend.Join when correcting offset for student
It examines.Loudspeaker plays the role of reminding student in vehicle shift ideal path.When the driving path of vehicle deviates ideal to the left
When path, loudspeaker issues voice prompting student " vehicle deviates to the left "；When vehicle deviates to the right when ideal path, loudspeaker
It issues voice prompting student " vehicle deviates to the right ".
Further, for convenience of driver's operation, can to offset given threshold, when offset is more than the threshold value,
Loudspeaker just starts working, and prompts driver.
The step of being related in above-described embodiment and algorithm are equally applicable to the sundry item training of subject two.
Embodiment 2
Referring to Fig. 1, the present embodiment provide a kind of automobile self training device based on strap-down navigation and area array cameras and its
Algorithm, device include 2 area array cameras, interior client, liquid crystal display and loudspeaker.Wherein:
Referring to fig. 2,2 area array cameras are respectively arranged in the two sides that vehicle leans on near-car head, fall for starting practice in student
When library or side are stopped, the relative tertiary location relationship between vehicle and starting point is detected to obtain the initial position of vehicle and initially
Yaw angle.A/D conversion module, 2.4G radio-frequency module and MCU module are integrated in area array cameras.
2.4G radio-frequency module, MCU module and digital strap-down navigation device are integrated in interior client, for falling in student
When library or side parking practice, the location information of vehicle is obtained in real time and the relative position for calculating vehicle and ideal path is closed
System, and error signal is passed through into display screen or loudspeaker feedback to student.
Liquid crystal display is used to show the offset of vehicle, while the progress bar on liquid crystal display shows the adjustment progress of student.
Loudspeaker plays suggesting effect, and when vehicle deviates ideal path to the left or to the right, loudspeaker can be provided accordingly
Prompt.
Interior client includes 2.4G radio-frequency module, MCU module and digital strap-down navigation device, knot in the present embodiment
Composition is as shown in Figure 1.
Claims (10)
1. a kind of automobile self training method based on strap-down navigation and area array cameras drives the study of Self-help vehicle formula for student
Library or side parking, which comprises the following steps:
Step 1,2 area array cameras are separately mounted to the two sides of vehicle headstock, when student starts practice library or side parking
When, the relative positional relationship between vehicle and library or side parking start line is detected by 2 area array cameras to obtain vehicle
Initial position and initial yaw angle；
Step 2, the coordinate at each turning point that vehicle needs to turn to during library or lateral parking is determined, according to
The vehicle initial position arrived and initial yaw angle carry out curve fitting to ideal planning driving path using interpolation method；
Step 3, real time position and yaw angle of the vehicle during library or lateral parking are obtained；
Step 4, the vehicle centroid position coordinates (X fallen according to vehicle during library or side coil(w), Y(w)) and vehicle body yaw angle theta,
The ideal path normal online to vehicle body yaw angle theta is acquired, which is vehicle location offset d；
Step 5, display vehicle location offset d is carried out by being mounted on the display screen before driver's cabin, if student adjusts vehicle
It is in the right direction, then show that the vehicle location offset d of screen display reduces；If student adjusts the anisotropy of vehicle, show
The vehicle location offset d of screen display increases, and driving Self-help vehicle formula to reach student falls library or side parking；
Vehicle cab is also equipped with loudspeaker:
If the driving path of vehicle deviates ideal path to the left, loudspeaker issues voice " vehicle deviates to the left "；
If the driving path of vehicle deviates to the right ideal path, loudspeaker issues voice " vehicle deviates to the right ".
2. automobile self training method as described in claim 1, which is characterized in that obtain the first of vehicle described in step 1
Beginning position and initial yaw angle include:
Step 11, by 2 area array cameras with shooting garages reticle image, using the ground reticle image as original image, and it is right
Original image carries out noise reduction pretreatment, obtains pretreatment image；
Step 12, the color space of pretreatment image is converted into HSV by BGR, and binary conversion treatment is carried out to pretreatment image,
Obtain binary image；
Step 13, after to binary image progress edge extracting, the straight line after detection edge extracting in bianry image, and extract
The coordinate of the pixel is fitted by the pixel at straight line, with obtaining garage slope of the graticule in image coordinate system
k1,k2With intercept d1,d2；
Step 14, the initial position by formula 1, formula 2 and formula 3 calculating vehicle and initial yaw angle, i.e. Vehicular yaw angle θ0And vehicle
The horizontal distance D of central point and sideline1,D2:
In formula, aij、bij、cij, m, n, q be calibration coefficient.
3. automobile self training method as described in claim 1, which is characterized in that interpolation method described in step 2 is specially
Cubic spline interpolation.
4. automobile self training method as described in claim 1, which is characterized in that acquisition vehicle described in step 3 is falling
Real time position and yaw angle during library or lateral parking include:
Step 31, the real-time angular velocity signal ω of vehicle is obtained, and ω is filtered；
Step 32, the real-time yaw angle theta of vehicle is calculated by formula 4:
θt=θt-1+∫ωdt+θ0Formula 4
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., ω are the real-time angular speed of vehicle
Signal；
Step 33, the real-time acceleration signal a of vehicle is acquired(B), and to a(B)It is filtered；
Step 34, coordinate X of the vehicle centroid under earth coordinates is calculated by formula 5 and formula 6(w)And Y(w):
X(w)=∫ ∫ (a(B)cosθdt)+X0Formula 5
Y(w)=∫ ∫ (a(B)sinθdt)+Y0Formula 6
Wherein, X(w)And Y(w)Respectively coordinate of the vehicle centroid under earth coordinates, θ are the real-time yaw angle of vehicle, X0、Y0
The respectively initial coordinate of vehicle centroid.
5. automobile self training method as claimed in claim 4, which is characterized in that by formula 7 to the ω described in step 31 into
Row filtering processing:
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., K are filter factor,When for t
Carve angular velocity signal estimated value.
6. a kind of automobile self training device based on strap-down navigation and area array cameras drives the study of Self-help vehicle formula for student
Library or side parking, which is characterized in that including image capture module, default ideal path module, obtain module, acquisition in real time
Offset module, self-service driving module；
Described image acquisition module includes 2 area array cameras, and 2 area array cameras are separately mounted to two that vehicle leans on near-car head
Side, for detecting vehicles by 2 area array cameras and by 2 faces gust phases when student starts to practice falling library or side parking
Machine detection vehicle and library or side parking start line between relative positional relationship with obtain vehicle initial position and initially
Yaw angle；
The default ideal path module is for determining that vehicle needs to turn to each turn during library or lateral parking
To at coordinate ideal planning driving path is carried out using interpolation method according to obtained vehicle initial position and initial yaw angle
Curve matching；
The real-time acquisition module is for obtaining real time position and yaw angle of the vehicle during library or lateral parking；
Vehicle centroid position coordinates (the X for obtaining offset module and being used to be fallen according to vehicle during library or side coil(w), Y(w)) and vehicle body yaw angle theta, the ideal path normal online to vehicle body yaw angle theta is acquired, which is vehicle position
Set offset d；
The self-service driving module includes display screen and loudspeaker, carries out display vehicle by being mounted on the display screen before driver's cabin
Position offset d shows that the vehicle location offset d of screen display reduces if student adjusts the in the right direction of vehicle；If learning
The anisotropy of member's adjustment vehicle then shows that the vehicle location offset d of screen display increases, and drives vehicle certainly to reach student
Help formula fall library or side parking；
Vehicle cab is also equipped with loudspeaker:
If the driving path of vehicle deviates ideal path to the left, loudspeaker issues voice " vehicle deviates to the left "；
If the driving path of vehicle deviates to the right ideal path, loudspeaker issues voice " vehicle deviates to the right ".
7. automobile self training device as claimed in claim 6, which is characterized in that acquisition vehicle described in image capture module
Initial position and initial yaw angle include:
Step 11, by 2 area array cameras with shooting garages reticle image, using the ground reticle image as original image, and it is right
Original image carries out noise reduction pretreatment, obtains pretreatment image；
Step 12, the color space of pretreatment image is converted into HSV by BGR, and binary conversion treatment is carried out to pretreatment image,
Obtain binary image；
Step 13, straight in image after detection edge extracting after carrying out edge extracting to binary image using Caany operator
Line, and the pixel at straight line is extracted, it is fitted according to the coordinate of the pixel, graticule is sat in image with obtaining garage
Slope k in mark system1,k2With intercept d1,d2；
Step 14, the initial position by formula 1, formula 2 and formula 3 calculating vehicle and initial yaw angle, i.e. Vehicular yaw angle θ0And vehicle
The horizontal distance D of central point and sideline1,D2:
In formula, aij、bij、cij, m, n, q be calibration coefficient.
8. automobile self training device as claimed in claim 6, which is characterized in that inserted described in default ideal path module
Value method is specially cubic spline interpolation.
9. automobile self training device as claimed in claim 6, which is characterized in that obtain acquisition vehicle described in module in real time
Real time position and yaw angle during library or lateral parking include:
Step 31, the real-time angular velocity signal ω of vehicle is obtained, and ω is filtered；
Step 32, the real-time yaw angle theta of vehicle is calculated by formula 4:
θt=θt-1+∫ωdt+θ0Formula 4
Wherein, t is value at the time of vehicle falls during library or lateral parking, and t=1,2 ..., ω are the real-time angular speed of vehicle
Signal；
Step 33, the real-time acceleration signal a of vehicle is acquired(B), and to a(B)It is filtered；
Step 34, coordinate X of the vehicle centroid under earth coordinates is calculated by formula 5 and formula 6(w)And Y(w):
X(w)=∫ ∫ (a(B)cosθdt)+X0Formula 5
Y(w)=∫ ∫ (a(B)sinθdt)+Y0Formula 6
Wherein, X(w)And Y(w)Respectively coordinate of the vehicle centroid under earth coordinates, θ are the real-time yaw angle of vehicle, X0、Y0
The respectively initial coordinate of vehicle centroid.
10. automobile self training device as claimed in claim 9, which is characterized in that by formula 7 to the ω described in step 31 into
Row filtering processing:
Wherein, t is value at the time of vehicle falls in library or side docking process, and t=1,2 ..., K are filter factor,For t moment
Angular velocity signal estimated value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611040667.XA CN106601076B (en) | 2016-11-11 | 2016-11-11 | A kind of automobile self training device and method based on inertial navigation and area array cameras |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611040667.XA CN106601076B (en) | 2016-11-11 | 2016-11-11 | A kind of automobile self training device and method based on inertial navigation and area array cameras |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106601076A CN106601076A (en) | 2017-04-26 |
| CN106601076B true CN106601076B (en) | 2019-06-18 |
Family
ID=58592767
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611040667.XA Expired - Fee Related CN106601076B (en) | 2016-11-11 | 2016-11-11 | A kind of automobile self training device and method based on inertial navigation and area array cameras |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106601076B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112486156B (en) * | 2019-09-10 | 2023-09-19 | 中车株洲电力机车研究所有限公司 | Automatic tracking control system and control method for vehicle |
| CN111645688B (en) * | 2019-09-19 | 2022-05-17 | 摩登汽车有限公司 | Method and device for calculating vehicle weight and gravity center in real time, electronic equipment and vehicle |
| CN113561897B (en) * | 2021-07-15 | 2022-08-12 | 河北三国新能源科技有限公司 | Method and system for judging ramp parking position of driving test vehicle based on panoramic all-round view |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1522892A (en) * | 2003-01-29 | 2004-08-25 | ��ʽ��������Զ�֯�������� | Parking assisting device |
| CN1591256A (en) * | 2003-08-28 | 2005-03-09 | 丰田自动车株式会社 | Vehicle backward movement assist device and vehicle parking assist device |
| CN1737501A (en) * | 2004-08-20 | 2006-02-22 | 爱信精机株式会社 | Parking auxiliary device for vehicle and parking auxiliary method |
| CN1987357A (en) * | 2006-12-26 | 2007-06-27 | 浙江工业大学 | Intelligent parking auxiliary device based on omnibearing computer sight |
| CN101300157A (en) * | 2005-10-31 | 2008-11-05 | 丰田自动车株式会社 | Parking assisting system |
| CN101635839A (en) * | 2008-07-25 | 2010-01-27 | 日产自动车株式会社 | Parking assistance apparatus and parking assistance method |
| JP2010195265A (en) * | 2009-02-26 | 2010-09-09 | Toyota Motor Corp | Parking facility |
| CN101945256A (en) * | 2009-07-01 | 2011-01-12 | 福特全球技术公司 | The auxiliary back of reversing camera system with touch-screen display |
| JP2011255812A (en) * | 2010-06-10 | 2011-12-22 | Aisin Seiki Co Ltd | Parking support device |
| CN103235327A (en) * | 2013-04-07 | 2013-08-07 | 清华大学 | GNSS/MINS (global navigation satellite system/micro-electro-mechanical systems inertial navigation system) super-deep combination navigation method, system and device |
| CN103499350A (en) * | 2013-09-28 | 2014-01-08 | 长安大学 | High-precision vehicle positioning method for fusing multi-source information under GPS (global positioning system) blind area and device |
| CN103591955A (en) * | 2013-11-21 | 2014-02-19 | 西安中科光电精密工程有限公司 | Combined navigation system |
| CN103847733A (en) * | 2012-11-28 | 2014-06-11 | 伊莱比特汽车公司 | Technique and apparatus for assisting driving a vehicle |
| CN103900611A (en) * | 2014-03-28 | 2014-07-02 | 北京航空航天大学 | Method for aligning two composite positions with high accuracy and calibrating error of inertial navigation astronomy |
| CN104197944A (en) * | 2014-03-03 | 2014-12-10 | 明日交通公司 | Position sensing system of intelligent vehicle navigation |
| CN104608692A (en) * | 2013-11-05 | 2015-05-13 | 现代摩比斯株式会社 | Parking auxiliary system and the method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010064546A (en) * | 2008-09-09 | 2010-03-25 | Toyota Industries Corp | Parking assist device |
-
2016
- 2016-11-11 CN CN201611040667.XA patent/CN106601076B/en not_active Expired - Fee Related
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1522892A (en) * | 2003-01-29 | 2004-08-25 | ��ʽ��������Զ�֯�������� | Parking assisting device |
| CN1591256A (en) * | 2003-08-28 | 2005-03-09 | 丰田自动车株式会社 | Vehicle backward movement assist device and vehicle parking assist device |
| CN1737501A (en) * | 2004-08-20 | 2006-02-22 | 爱信精机株式会社 | Parking auxiliary device for vehicle and parking auxiliary method |
| CN101300157A (en) * | 2005-10-31 | 2008-11-05 | 丰田自动车株式会社 | Parking assisting system |
| CN1987357A (en) * | 2006-12-26 | 2007-06-27 | 浙江工业大学 | Intelligent parking auxiliary device based on omnibearing computer sight |
| CN101635839A (en) * | 2008-07-25 | 2010-01-27 | 日产自动车株式会社 | Parking assistance apparatus and parking assistance method |
| JP2010195265A (en) * | 2009-02-26 | 2010-09-09 | Toyota Motor Corp | Parking facility |
| CN101945256A (en) * | 2009-07-01 | 2011-01-12 | 福特全球技术公司 | The auxiliary back of reversing camera system with touch-screen display |
| JP2011255812A (en) * | 2010-06-10 | 2011-12-22 | Aisin Seiki Co Ltd | Parking support device |
| CN103847733A (en) * | 2012-11-28 | 2014-06-11 | 伊莱比特汽车公司 | Technique and apparatus for assisting driving a vehicle |
| CN103235327A (en) * | 2013-04-07 | 2013-08-07 | 清华大学 | GNSS/MINS (global navigation satellite system/micro-electro-mechanical systems inertial navigation system) super-deep combination navigation method, system and device |
| CN103499350A (en) * | 2013-09-28 | 2014-01-08 | 长安大学 | High-precision vehicle positioning method for fusing multi-source information under GPS (global positioning system) blind area and device |
| CN104608692A (en) * | 2013-11-05 | 2015-05-13 | 现代摩比斯株式会社 | Parking auxiliary system and the method |
| CN103591955A (en) * | 2013-11-21 | 2014-02-19 | 西安中科光电精密工程有限公司 | Combined navigation system |
| CN104197944A (en) * | 2014-03-03 | 2014-12-10 | 明日交通公司 | Position sensing system of intelligent vehicle navigation |
| CN103900611A (en) * | 2014-03-28 | 2014-07-02 | 北京航空航天大学 | Method for aligning two composite positions with high accuracy and calibrating error of inertial navigation astronomy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106601076A (en) | 2017-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106256606B (en) | A kind of lane departure warning method based on vehicle-mounted binocular camera | |
| CN106379237B (en) | Vehicle lane-changing overall process DAS (Driver Assistant System) based on augmented reality | |
| CN107554430A (en) | Vehicle blind zone view method, apparatus, terminal, system and vehicle | |
| CN109649384A (en) | A kind of parking assistance method | |
| CN110745140B (en) | Vehicle lane change early warning method based on continuous image constraint pose estimation | |
| CN107577996A (en) | A kind of recognition methods of vehicle drive path offset and system | |
| DE102009008113B4 (en) | Method and device for detecting a target parking position using two reference points and a parking assistance system using the same | |
| CN104751151B (en) | A kind of identification of multilane in real time and tracking | |
| CN106156723B (en) | A kind of crossing fine positioning method of view-based access control model | |
| CN107590438A (en) | A kind of intelligent auxiliary driving method and system | |
| CN107750364A (en) | Detected using the road vertically profiling of stable coordinate system | |
| CN106295560A (en) | The track keeping method controlled based on vehicle-mounted binocular camera and stagewise PID | |
| CN107462223A (en) | Driving sight distance self-operated measuring unit and measuring method before a kind of highway is turned | |
| CN105069859B (en) | Vehicle running state monitoring method and device | |
| DE102015105760B4 (en) | System for parking a vehicle | |
| CN107133985A (en) | A kind of vehicle-mounted vidicon automatic calibration method for the point that disappeared based on lane line | |
| CN105976402A (en) | Real scale obtaining method of monocular vision odometer | |
| CN106601076B (en) | A kind of automobile self training device and method based on inertial navigation and area array cameras | |
| CN107229908A (en) | A kind of method for detecting lane lines | |
| CN107284455B (en) | A kind of ADAS system based on image procossing | |
| DE102017128619A1 (en) | VEHICLE CONTROL USING ROAD ANGLE DATA | |
| DE10323915A1 (en) | Camera-based position detection for a road vehicle | |
| JP5310027B2 (en) | Lane recognition device and lane recognition method | |
| DE112015004011T5 (en) | VEHICLE IMAGING DEVICE | |
| EP3935560A1 (en) | Method for training and using a neural network to detect ego part position |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190618 Termination date: 20191111 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |