CN110458893A - The roll angle scaling method and system of advanced auxiliary driving vision detecting sensor - Google Patents
The roll angle scaling method and system of advanced auxiliary driving vision detecting sensor Download PDFInfo
- Publication number
- CN110458893A CN110458893A CN201910691585.9A CN201910691585A CN110458893A CN 110458893 A CN110458893 A CN 110458893A CN 201910691585 A CN201910691585 A CN 201910691585A CN 110458893 A CN110458893 A CN 110458893A
- Authority
- CN
- China
- Prior art keywords
- point
- identification point
- optimal identification
- roll angle
- optimal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
Abstract
The embodiment of the present invention provides the roll angle scaling method and system of a kind of advanced auxiliary driving vision detecting sensor, this method comprises: the posture information and the pixel translation distance in spotting object central projection to the sensor devices of camera that obtain camera;Based on pixel translation distance, the picture element position information of the identification point of spotting object is calculated by convolution kernel;Picture element position information is screened, candidate point aggregation zone is obtained;The optimization of sub-pix rank is carried out to candidate point aggregation zone, obtains the location information of optimal identification point;Wherein, optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;Roll angle is obtained by the location information of the optimal identification point of the optimal identification point in left side and right side.The embodiment of the present invention sets specific operating environment, and externally location information has carried out the other thick calibration of Centimeter Level in ginseng, and then externally roll angle carries out Fine calibration in ginseng, reaches the other stated accuracy of sub-pixel.
Description
Technical field
The present invention relates to driving technology fields, more particularly, to a kind of advanced auxiliary driving vision detecting sensor
Roll angle scaling method and system.
Background technique
Currently, more and more new passenger cars load advanced DAS (Driver Assistant System).According in Chinese automobile technical research
Heart standard and Nissan Motor joint publication " advanced driving assistance system (ADAS) research on standard map out a route figure research report
Accuse ", it is expected that 2015-2020, China carries and drives auxiliary (DA), the vehicle market occupation rate of part automatic Pilot (PA) is about
50%;It is about 15% that 2020-2025, DA, PA vehicle occupancy rate, which keep stable, highly automated driving (HA) vehicle occupancy rate,;
To 2025-2030, fully automated driving (FA) vehicle market occupation rate is close to 10%.
Advanced auxiliary drives, the function of fully automated driving is largely related to visual sensor, even with visual sensing
Based on device.Therefore, the reliability of visual sensor and safety are particularly important.Wherein, the calibration of visual sensor refers to
It is system data two-dimensional in visual sensor, is converted to the calibration that can carry out the three-dimensional data of vehicle action decision judgement
Process.Advanced DAS (Driver Assistant System), fully automated control loop can all carry out this calibration operation before incoming terminal user, with
Ensure the reliability and safety of system.
General calibration system includes outer ginseng calibration and internal reference calibration, and outer ginseng calibration includes location information and posture information again
Calibration.In the driving of advanced auxiliary and fully automated driving field, the internal reference of general visual sensor has been demarcated by manufacturer to be mentioned
For what is demarcated in general vehicle-mounted vision system is then outer ginseng.Wherein, it is outer ginseng location information according to onboard system design position into
Row installation, accounts for smaller influence in calibration in accuracy.On the other hand, posture information accounts for larger shadow in accuracy in outer ginseng
It rings.
In existing technology, the calibration for externally joining posture information is often concerned with its yaw angle (Yaw) and pitch angle
(Pitch) calibration, and have ignored the calibration to roll angle (Roll).The result is that measurement should be roll caused by this way
The value at angle is to yaw angle and pitch angle has been projected, so as to cause the decline of a certain amount of reliability and safety.Recently, with
Hardware aspect is substantially improved with visual sensor resolution ratio, and such as by, VGA to 720p, 1080p, 2k etc. become master in recent years
Stream, the prospect of following 4K, 8K provide the foundation to technology upgrading.One side application function needs to extend, as vision positions in real time
The drafting of system and high-precision map, the tracking system of accurate Pedestrians and vehicles.The security level driven is finally assisted to need to be promoted,
Such as automatically kept (LKA) from lane departure warning (LDW) to current road, advanced auxiliary drives function from towards driver
Prompt risen to the direct vehicle control towards Body Control, the requirement to vision sensor calibration is then further stringent.
So then becoming further urgent for the calibration demand of accurate roll angle in in-vehicle camera calibration.
Summary of the invention
To solve the above-mentioned problems, the embodiment of the present invention provides one kind and overcomes the above problem or at least be partially solved
State the roll angle scaling method and system of the advanced auxiliary driving vision detecting sensor of problem.
According to a first aspect of the embodiments of the present invention, a kind of roll angle of advanced auxiliary driving vision detecting sensor is provided
Scaling method, this method comprises: obtain camera posture information and the central projection of spotting object to camera sensor devices
On pixel translation distance;Based on pixel translation distance, the pixel position of the identification point of spotting object is calculated by convolution kernel
Confidence breath;Picture element position information is screened, candidate point aggregation zone is obtained;Sub-pixel is carried out to candidate point aggregation zone
Do not optimize, obtains the location information of optimal identification point;Wherein, optimal identification point includes the optimal mark of the optimal identification point in left side and right side
Know point;Roll angle is obtained by the location information of the optimal identification point of the optimal identification point in left side and right side.
Second aspect according to embodiments of the present invention provides a kind of roll angle of advanced auxiliary driving vision detecting sensor
Calibration system, the system include: acquisition module, and the posture information and the central projection of spotting object for obtaining camera are to phase
Pixel translation distance on the sensor devices of machine;Computing module calculates bid by convolution kernel for being based on pixel translation distance
Set the goal object identification point picture element position information;Screening module obtains candidate point for screening to picture element position information
Aggregation zone;Optimization module obtains the position of optimal identification point for carrying out the optimization of sub-pix rank to candidate point aggregation zone
Information;Wherein, optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;Module is obtained, for passing through left side
The location information of the optimal identification point of optimal identification point and right side obtains roll angle.
According to a third aspect of the embodiments of the present invention, a kind of electronic equipment, including memory, processor and storage are provided
On a memory and the computer program that can run on a processor, processor is realized various such as first aspect when executing program
The roll of advanced auxiliary driving vision detecting sensor provided by any possible implementation in possible implementation
Angle scaling method.
According to a fourth aspect of the embodiments of the present invention, a kind of non-transient computer readable storage medium is provided, is deposited thereon
Computer program is contained, is realized in the various possible implementations such as first aspect when which is executed by processor
The roll angle scaling method of advanced auxiliary driving vision detecting sensor provided by any possible implementation.
The roll angle scaling method and system of advanced auxiliary driving vision detecting sensor provided in an embodiment of the present invention, if
Specific operating environment is determined, externally location information has carried out the other thick calibration of Centimeter Level in ginseng, then roll angle in external ginseng
Fine calibration is carried out, the other stated accuracy of sub-pixel is reached.On testing vision sensor, positioned at the barrier of 50m distance,
The vision transformed error of position is then dropped within 2cm by 8cm.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described.It should be evident that the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these figures.
Fig. 1 is the stream of the roll angle scaling method of advanced auxiliary driving vision detecting sensor provided in an embodiment of the present invention
Journey schematic diagram;
Fig. 2 is that in-vehicle camera provided in an embodiment of the present invention demarcates environment set schematic diagram;
Fig. 3 is in-vehicle camera image-forming principle schematic diagram provided in an embodiment of the present invention;
Fig. 4 is A point provided in an embodiment of the present invention and B point schematic diagram;
Fig. 5 is A point convolution kernel schematic diagram provided in an embodiment of the present invention;
Fig. 6 is B point convolution kernel schematic diagram provided in an embodiment of the present invention;
Fig. 7 is filtered candidate point area schematic provided in an embodiment of the present invention;
Fig. 8 is that the value provided in an embodiment of the present invention using maximum value nearby coordinates calculates hump schematic diagram;
Fig. 9 is the flow diagram provided in an embodiment of the present invention that hump method is calculated using centre-of-gravity principle;
Figure 10 is that the hump provided in an embodiment of the present invention for calculating hump using centre-of-gravity principle calculates schematic diagram;
Figure 11 is that provided in an embodiment of the present invention other calculate the calculating of quadratic equation with one unknown model in the model of hump
Schematic diagram;
Figure 12 is that provided in an embodiment of the present invention other calculate the just too calculating signal of distributed model in the model of hump
Figure;
Figure 13 is that roll angle provided in an embodiment of the present invention calculates schematic diagram;
Figure 14 is the roll angle calibration system of advanced auxiliary driving vision detecting sensor provided in an embodiment of the present invention
Structural schematic diagram;
Figure 15 is the entity structure schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical solution in the embodiment of the present invention is explicitly described, it is clear that described embodiment is the present invention
A part of the embodiment, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not having
Every other embodiment obtained under the premise of creative work is made, shall fall within the protection scope of the present invention.
The embodiment of the present invention belongs to advanced auxiliary and drives and automatic Pilot technical field, especially a kind of visual perception sensing
The adjustment automation of angle is rotated in device external parameter, improves the scaling method of its accuracy, and in particular to advanced auxiliary driving,
The application of parameter configuration precision of the automatic Pilot technology before Driving Scene landing is driven with meeting corresponding advanced auxiliary, is automatic
The reliability and safety of driving technology.
Referring to Fig. 1, the embodiment of the present invention provides a kind of roll angle calibration side of advanced auxiliary driving vision detecting sensor
Method, comprising:
Step 101, according to scene is preset, the translation gesture information for obtaining camera and the central projection of spotting object are to phase
Pixel translation distance on the sensor devices of machine.
As a kind of alternative embodiment, the translation gesture information of camera is obtained, comprising: operation ring is carried out according to default scene
Border configuration sets the goal object in operating environment acceptance of the bid;It is obtained according to the location information of the location information of spotting object and camera
Obtain the translation gesture information of camera.
Wherein, referring to fig. 2, operating environment arrangement, spotting object in environment are carried out according to design scenario, camera position is
It is known that solving posture information.Specifically, operating environment configuration is carried out.Vehicle centre-line extends to the center of scaling board.Scaling board
Location information and the location information of vehicle be known.Camera is installed in front point in the car close to vehicle center line position, phase
Translational shifting of the machine apart from vehicle center position is set to it is known that trueness error is millimeter rank.The rotation attitude of camera be it is unknown, i.e.,
The objective result of calculating.
It is calculated in object central projection to camera sensor devices referring to Fig. 3 according to CMOS camera pinhole imaging system principle
Pixel translation distance.Specifically, it is determined that the cross of imaging center, ordinate I=(Ix,Iy).Internal optical axis of this coordinate in camera
On line.The point distance that imaging center distance template central point is incident upon on picture can use Fig. 3 and be calculated, i.e. offset T
=(Tx,Ty).The coordinate that then target's center projects on picture is C=I+T.Wherein, focal length (pixel) can be joined by camera
Number obtains.
Step 102, the form Design based on object, the identification point of spotting object is calculated by corresponding convolution kernel
Picture element position information.
As a kind of alternative embodiment, the form Design based on object calculates calibration mesh by corresponding convolution kernel
Mark object identification point picture element position information, comprising: is obtained in the imaging of spotting object by different convolution kernels A point with
B point;Wherein, A point is for determining that identification point, B point are used to determine the correctness of A point.
As a kind of alternative embodiment, A point corresponds to the first kernel for using 19*19, and upper left and bottom right are in the first kernel
The numerical value of 8*8 is all -1 matrix, and upper right and lower-left are that the numerical value of 8*8 is all 1 matrix, other positions are 0;B point correspondence makes
With the second kernel of 9*9, upper left and bottom right are that the numerical value of 3*3 is all 1 matrix in the second kernel, and upper right and lower-left are 3*3's
Numerical value is all -1 matrix.
Specifically, referring to fig. 4,5 and 6, the picture element position information of object or so identification point is calculated using different convolution.
A, B point are found in imaging.A is calculated by two-dimensional convolution, and it is 19*19 kernel, upper left and bottom right in kernel that convolution kernels, which use,
- 1 matrix is all for the numerical value of 8*8, upper right and lower-left are that the numerical value of 8*8 is all 1 matrix in kernel, other positions are 0.Phase
The convolution kernels of the B answered are 9*9, and upper left and bottom right are the matrix that 3*3 is all 1, and upper right and lower-left are the matrix that 3*3 is all -1.
Show that A, B greater than threshold value are combined, A point searching accurate location, then the B point of the point right A or left determines A point
Correctness.
Step 103 screens picture element position information, obtains candidate point aggregation zone.
As a kind of alternative embodiment, picture element position information is screened, after obtaining candidate point aggregation zone, is also wrapped
It includes: picture element position information is screened, form candidate point after the point more than preset threshold is collected;And assemble candidate point
Region is marked candidate point aggregation zone as candidate point aggregation zone, using black surround.
Specifically, referring to Fig. 7, candidate point is formed after the point that convolution kernel calculated result is more than preset threshold is collected, they
It will be used for subsequent calculating, candidate point aggregation zone to be marked using black surround.In other words, convolution kernel result is screened, is more than pre-
If this of threshold value is the alternative point for the scheme that optimizes, these points are marked using black surround.
Step 104 carries out the optimization of sub-pix rank to candidate point aggregation zone, obtains the location information of optimal identification point;
Wherein, optimal identification point includes the optimal identification point of the optimal identification point in left side and right side.
Referring to Fig. 8, the optimization of sub-pix rank is carried out to candidate point aggregation zone using optimization algorithm, obtains optimal mark
Point.The accurate location of A point is properly termed as hump coordinate.Its source is the region of black surround mark in Fig. 7.The embodiment of the present invention is logical
It crosses and finds the changing rules of these points and determine hump coordinate.Therefore hump coordinate precision has been increased to sub-pix rank.
Or more points are selected at 5 points according to actual operation demand and accuracy requirement.
Referring to Fig. 9 and Figure 10,5 points of filterings based on center of gravity calculation are can be used in optimization algorithm;Referring to Figure 11 and Figure 12,
It is also possible to quadratic equation with one unknown curve or normal distribution song based on fitting.Also, according to actual operation demand and precision
Or more points may be selected at 5 points in demand.
As a kind of alternative embodiment, the optimization of sub-pix rank is carried out to candidate point aggregation zone, obtains optimal identification point
Location information, comprising: 5 filter methods based on center of gravity, to A point carry out the other optimization of sub-pixel;
Wherein, the process of optimization is to obtain the filter result progress hump coordinate calculating of five points closed on, is obtained most
The location information of excellent identification point;Wherein, five points closed on specifically include five points of the relative position [- 2,2] in abscissa, with
And ordinate is in the point of proximity of relative position [- 2,2];Wherein, the accurate location of A point is hump coordinate.
Specifically, referring to Fig. 9 and Figure 10, a kind of five based on center of gravity filter method is described, realizes and Asia is carried out to A point
The optimization of pixel scale.The process of optimization is that the filter result progress hump for five points that acquirement closes on must calculate.It closes on
Five points specifically include five points and ordinate facing in relative position [- 2,2] of the relative position [- 2,2] in abscissa
Near point.
As a kind of alternative embodiment, the optimization of sub-pix rank is carried out to candidate point aggregation zone, obtains optimal identification point
Location information, comprising: be based on curve-fitting method, to A point carry out the other optimization of sub-pixel;
Wherein, the process of optimization is the variation using the model of quadratic equation with one unknown or the model of normal distribution to A point
Rule is fitted;After formula after seeking fitting, the coordinate of maximum value, the position letter of as optimal identification point are sought to formula
Breath.
Specifically, referring to Figure 11 and Figure 12, other other optimization methods of sub-pixel based on curve matching are described.
The changing rule of these points is fitted using the model of quadratic equation with one unknown or the model of normal distribution.After seeking fitting
Then formula seeks the coordinate of maximum value, the coordinate of as A point to formula.
Step 105 obtains roll angle by the location information of the optimal identification point of the optimal identification point in left side and right side.
Specifically, referring to Figure 13, roll angle posture information is calculated according to the optimal identification point in left and right.Left and right A is found out respectively
After the coordinate of point, carries out last roll angle and seek.After left side A point and right side A point line and horizontal sextant angle is roll angle:
To sum up, the scaling method that the embodiment of the present invention proposes is demarcated applied to outer ginseng, and common outer parameter can base according to extraction
In the internal reference data that visual sensor manufacturer provides, outer parameter evidence is calculated by a series of scaling methods.The present invention is implemented
Example sets specific operating environment, and externally location information has carried out the other thick calibration of Centimeter Level in ginseng, then externally horizontal in ginseng
Roll angle carries out Fine calibration, reaches the other stated accuracy of sub-pixel.On testing vision sensor, positioned at the barrier of 50m distance
Hinder object, the vision transformed error of position is then dropped within 2cm by 8cm.Note: visual sensor resolution ratio is 1920*1080, burnt
Away from 3.6395mm.
Content based on the above embodiment, the embodiment of the invention provides a kind of advanced auxiliary driving vision detecting sensors
Roll angle calibration system, roll angle calibration system of the advanced auxiliary driving vision detecting sensor is for executing the above method
The roll angle scaling method of advanced auxiliary driving vision detecting sensor in embodiment.Referring to Figure 14, which includes: to obtain
Module 201, the pixel on posture information and spotting object central projection to the sensor devices of camera for obtaining camera
Translation distance;Computing module 202 calculates the identification point of spotting object by convolution kernel for being based on pixel translation distance
Picture element position information;Screening module 203 obtains candidate point aggregation zone for screening to picture element position information;Optimization
Module 204 obtains the location information of optimal identification point for carrying out the optimization of sub-pix rank to candidate point aggregation zone;Wherein,
Optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;Module 205 is obtained, for passing through the optimal mark in left side
The location information of point and the optimal identification point in right side obtains roll angle.
The embodiment of the invention provides a kind of electronic equipment, and as shown in figure 15, which includes: processor
(processor) 501, communication interface (Communications Interface) 502, memory (memory) 503 and communication
Bus 504, wherein processor 501, communication interface 502, memory 503 complete mutual communication by communication bus 504.
Processor 501 can call the computer program that can be run on memory 503 and on processor 501, to execute above-mentioned each reality
Apply the roll angle scaling method of the advanced auxiliary driving vision detecting sensor of example offer, for example, obtain the posture of camera
Pixel translation distance in information and spotting object central projection to the sensor devices of camera;Based on pixel translation distance,
The picture element position information of the identification point of spotting object is calculated by convolution kernel;Picture element position information is screened, is obtained
Candidate point aggregation zone;The optimization of sub-pix rank is carried out to candidate point aggregation zone, obtains the location information of optimal identification point;Its
In, optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;It is optimal by the optimal identification point in left side and right side
The location information of identification point obtains roll angle.
In addition, the logical order in above-mentioned memory 503 can be realized by way of SFU software functional unit and conduct
Independent product when selling or using, can store in a computer readable storage medium.Based on this understanding, originally
Substantially the part of the part that contributes to existing technology or the technical solution can be in other words for the technical solution of invention
The form of software product embodies, which is stored in a storage medium, including some instructions to
So that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation of the present invention
The all or part of the steps of example method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-
Only Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can be with
Store the medium of program code.
The embodiment of the present invention also provides a kind of non-transient computer readable storage medium, is stored thereon with computer program,
The computer program is implemented to carry out the various embodiments described above offer advanced auxiliary driving vision perception when being executed by processor passes
The roll angle scaling method of sensor, for example, the posture information for obtaining camera and the central projection of spotting object are to camera
Sensor devices on pixel translation distance;Based on pixel translation distance, the mark of spotting object is calculated by convolution kernel
The picture element position information of point;Picture element position information is screened, candidate point aggregation zone is obtained;To candidate point aggregation zone into
The optimization of row sub-pix rank, obtains the location information of optimal identification point;Wherein, optimal identification point include the optimal identification point in left side and
The optimal identification point in right side;Roll angle is obtained by the location information of the optimal identification point of the optimal identification point in left side and right side.
The embodiments such as electronic equipment described above are only schematical, wherein unit as illustrated by the separation member
It may or may not be physically separated, component shown as a unit may or may not be physics list
Member, it can it is in one place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of the modules achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case where, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Certain Part Methods of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features;
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of roll angle scaling method of advanced auxiliary driving vision detecting sensor characterized by comprising
Obtain the posture information of camera and the pixel translation in spotting object central projection to the sensor devices of the camera
Distance;
Based on the pixel translation distance, believed by the location of pixels that convolution kernel calculates the identification point of the spotting object
Breath;
The picture element position information is screened, candidate point aggregation zone is obtained;
The optimization of sub-pix rank is carried out to the candidate point aggregation zone, obtains the location information of optimal identification point;Wherein, described
Optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;
Roll angle is obtained by the location information of the optimal identification point of the optimal identification point in the left side and the right side.
2. the method according to claim 1, wherein the posture information for obtaining camera, comprising:
Operating environment configuration is carried out according to default scene, the spotting object is demarcated in the operating environment;
The posture information of the camera is obtained according to the location information of the location information of the spotting object and the camera.
3. being calculated the method according to claim 1, wherein being based on the pixel translation distance by convolution kernel
The picture element position information of the identification point of the spotting object out, comprising:
A point and B point are obtained in the imaging of the spotting object by different convolution kernels;Wherein, described in A point is used to determine
Identification point, B point are used to determine the correctness of A point.
4. according to the method described in claim 3, it is characterized in that,
The A point corresponds to the first kernel for using 19*19, and the numerical value that upper left and bottom right are 8*8 in first kernel is all -1
Matrix, upper right and lower-left are the matrix that the numerical value of 8*8 is all 1, other positions are 0;
The B point corresponds to the second kernel for using 9*9, and upper left and bottom right are that the numerical value of 3*3 is all 1 square in second kernel
Battle array, upper right and lower-left are the matrix that the numerical value of 3*3 is all -1.
5. being obtained candidate the method according to claim 1, wherein being screened to the picture element position information
After point aggregation zone, further includes:
The picture element position information is screened, forms candidate point after the point more than preset threshold is collected;And by candidate point
The region of aggregation is marked the candidate point aggregation zone as the candidate point aggregation zone, using black surround.
6. according to the method described in claim 4, it is characterized in that, excellent to candidate point aggregation zone progress sub-pix rank
Change, obtain the location information of optimal identification point, comprising: it is other to carry out sub-pixel to A point for 5 filter methods based on center of gravity
Optimization;
Wherein, the process of optimization is to obtain the filter result progress hump coordinate calculating of five points closed on, obtains optimal mark
Know the location information of point;Wherein, five closed on the point includes five points of relative position [- 2,2] in abscissa, and vertical
Point of proximity of the coordinate in relative position [- 2,2];Wherein, the accurate location of A point is the hump coordinate.
7. according to the method described in claim 4, it is characterized in that, excellent to candidate point aggregation zone progress sub-pix rank
Change, obtain the location information of optimal identification point, comprising: is based on curve-fitting method, the other optimization of sub-pixel is carried out to A point;
Wherein, the process of optimization is the changing rule using the model of the model of quadratic equation with one unknown or normal distribution to A point
It is fitted;After formula after seeking fitting, the coordinate of maximum value, the position letter of the as described optimal identification point are sought to formula
Breath.
8. a kind of roll angle calibration system of advanced auxiliary driving vision detecting sensor characterized by comprising
Obtain module, for obtain camera posture information and the central projection of spotting object to the camera sensor devices
On pixel translation distance;
Computing module calculates the identification point of the spotting object by convolution kernel for being based on the pixel translation distance
Picture element position information;
Screening module obtains candidate point aggregation zone for screening to the picture element position information;
Optimization module obtains the position of optimal identification point for carrying out the optimization of sub-pix rank to the candidate point aggregation zone
Information;Wherein, the optimal identification point includes the optimal identification point of the optimal identification point in left side and right side;
Module is obtained, for being rolled by the location information of the optimal identification point of the optimal identification point in the left side and the right side
Angle.
9. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor is realized advanced auxiliary as described in any one of claim 1 to 7 when executing described program
The step of helping the roll angle scaling method of driving vision detecting sensor.
10. a kind of non-transient computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer
The cross of the advanced auxiliary driving vision detecting sensor as described in any one of claim 1 to 7 is realized when program is executed by processor
The step of roll angle scaling method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910691585.9A CN110458893B (en) | 2019-07-29 | 2019-07-29 | Roll angle calibration method and system for advanced driving assistance visual perception sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910691585.9A CN110458893B (en) | 2019-07-29 | 2019-07-29 | Roll angle calibration method and system for advanced driving assistance visual perception sensor |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110458893A true CN110458893A (en) | 2019-11-15 |
CN110458893B CN110458893B (en) | 2021-09-24 |
Family
ID=68483893
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910691585.9A Active CN110458893B (en) | 2019-07-29 | 2019-07-29 | Roll angle calibration method and system for advanced driving assistance visual perception sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110458893B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101793517A (en) * | 2010-02-09 | 2010-08-04 | 北京航空航天大学 | Online quick method for improving accuracy of attitude determination of airborne platform |
WO2015160287A1 (en) * | 2014-04-14 | 2015-10-22 | Saab Vricon Systems Ab | A method and system for estimating information related to a vehicle pitch and/or roll angle |
US20170134722A1 (en) * | 2013-05-21 | 2017-05-11 | Magna Electronics Inc. | Vehicle vision system with targetless camera calibration |
CN108648241A (en) * | 2018-05-17 | 2018-10-12 | 北京航空航天大学 | A kind of Pan/Tilt/Zoom camera field calibration and fixed-focus method |
EP3389015A1 (en) * | 2017-04-13 | 2018-10-17 | Continental Automotive GmbH | Roll angle calibration method and roll angle calibration device |
CN110057295A (en) * | 2019-04-08 | 2019-07-26 | 河海大学 | It is a kind of to exempt from the monocular vision plan range measurement method as control |
-
2019
- 2019-07-29 CN CN201910691585.9A patent/CN110458893B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101793517A (en) * | 2010-02-09 | 2010-08-04 | 北京航空航天大学 | Online quick method for improving accuracy of attitude determination of airborne platform |
US20170134722A1 (en) * | 2013-05-21 | 2017-05-11 | Magna Electronics Inc. | Vehicle vision system with targetless camera calibration |
WO2015160287A1 (en) * | 2014-04-14 | 2015-10-22 | Saab Vricon Systems Ab | A method and system for estimating information related to a vehicle pitch and/or roll angle |
EP3389015A1 (en) * | 2017-04-13 | 2018-10-17 | Continental Automotive GmbH | Roll angle calibration method and roll angle calibration device |
CN108648241A (en) * | 2018-05-17 | 2018-10-12 | 北京航空航天大学 | A kind of Pan/Tilt/Zoom camera field calibration and fixed-focus method |
CN110057295A (en) * | 2019-04-08 | 2019-07-26 | 河海大学 | It is a kind of to exempt from the monocular vision plan range measurement method as control |
Non-Patent Citations (2)
Title |
---|
ZACHARY TAYLOR 等: "Automatic Calibration of Lidar and Camera Images using Normalized Mutual Information", 《ROBOTICS AND AUTOMATION (ICRA)》 * |
向高林 等: "基于改进六位置法的一种MEMS加速度计标定补偿方案", 《重庆邮电大学学报( 自然科学版)》 * |
Also Published As
Publication number | Publication date |
---|---|
CN110458893B (en) | 2021-09-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107230218B (en) | Method and apparatus for generating confidence measures for estimates derived from images captured by vehicle-mounted cameras | |
WO2021004548A1 (en) | Vehicle speed intelligent measurement method based on binocular stereo vision system | |
US10317231B2 (en) | Top-down refinement in lane marking navigation | |
EP4016457A1 (en) | Positioning method and apparatus | |
CN109074490B (en) | Path detection method, related device and computer readable storage medium | |
CN112733812B (en) | Three-dimensional lane line detection method, device and storage medium | |
CN110203210A (en) | A kind of lane departure warning method, terminal device and storage medium | |
KR20210078530A (en) | Lane property detection method, device, electronic device and readable storage medium | |
JP6520740B2 (en) | Object detection method, object detection device, and program | |
CN110858405A (en) | Attitude estimation method, device and system of vehicle-mounted camera and electronic equipment | |
CN109849930B (en) | Method and device for calculating speed of adjacent vehicle of automatic driving automobile | |
CN109543493A (en) | A kind of detection method of lane line, device and electronic equipment | |
LU502288B1 (en) | Method and system for detecting position relation between vehicle and lane line, and storage medium | |
US11776277B2 (en) | Apparatus, method, and computer program for identifying state of object, and controller | |
CN112740225B (en) | Method and device for determining road surface elements | |
CN112862890B (en) | Road gradient prediction method, device and storage medium | |
CN112525147B (en) | Distance measurement method for automatic driving equipment and related device | |
CN107688174A (en) | A kind of image distance-finding method, system, storage medium and vehicle-mounted visually-perceptible equipment | |
WO2018149539A1 (en) | A method and apparatus for estimating a range of a moving object | |
CN113255444A (en) | Training method of image recognition model, image recognition method and device | |
CN114179788A (en) | Automatic parking method, system, computer readable storage medium and vehicle terminal | |
JP2021128705A (en) | Object state identification device | |
CN108629225A (en) | A kind of vehicle checking method based on several subgraphs and saliency analysis | |
CN112902911B (en) | Ranging method, device, equipment and storage medium based on monocular camera | |
CN112529011A (en) | Target detection method and related device |
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 |