CN109409311A - A kind of limit for height method for early warning based on binocular stereo vision - Google Patents
A kind of limit for height method for early warning based on binocular stereo vision Download PDFInfo
- Publication number
- CN109409311A CN109409311A CN201811316705.9A CN201811316705A CN109409311A CN 109409311 A CN109409311 A CN 109409311A CN 201811316705 A CN201811316705 A CN 201811316705A CN 109409311 A CN109409311 A CN 109409311A
- Authority
- CN
- China
- Prior art keywords
- coordinate system
- height
- world coordinate
- pixel
- axis
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
Abstract
The present invention relates to a kind of limit for height method for early warning based on binocular stereo vision, comprising: S1. obtains the disparity map of original image based on the original image that binocular camera is absorbed, and the current driving path of vehicle is obtained in world coordinate system;S2. it calculates each of disparity map pixel and projects to the first coordinate in world coordinate system;S3. judge that corresponding pixel is whether in driving path in disparity map according to the first coordinate, if so, the pixel is projected on vertical coordinate plane parallel with binocular camera image planes in world coordinate system;S4. barrier is judged whether there is based on preset height region on the vertical coordinate plane, and if it exists, then obtain the average height of barrier;S5. judgement and tracking that step S1 to S4 carries out multiframe to the barrier are repeated, further confirms that the barrier is limited and high indicate and issue early warning.According to the method for the present invention, calculation amount is small, and accuracy is high.
Description
Technical field
The present invention relates to a kind of limit for height method for early warning more particularly to a kind of pre- police of limit for height based on binocular stereo vision
Method.
Background technique
With the development of economy, Urban Transportation also flourishes therewith, in some roads of urban transportation, for limitation
Overheight Vehicles pass through, and often will increase limit for height barrier, and the generation for avoiding excessively high vehicle that Ka Qiao occurs, hits the accidents such as bridge is protected
Demonstrate,prove the normal pass of road.But such traffic accident still can still happen occasionally, currently, still without effective on vehicle
The device of accurate detection front limit for height barrier, needs driver independently to be observed, and often leads to driver in this way and estimates mistake
Miss and so on and cause an accident, but also be easy to happen driver carelessness the case where.
For example, application publication number be CN105915846A, it is entitled " it is a kind of list binocular multiplexing invader monitoring method and
The Chinese patent of system " discloses a kind of its technical solution of the monitoring method of invader are as follows: shooting background image;First video camera
Acquire the monitor video image of monitoring area;It is determined and is monitored according to the monitor video image that background image and the first video camera acquire
Whether region has invader entrance, then controls the starting of the second video camera in this way, step and this step before otherwise repeating;According to first
The video image of video camera shooting and the video image of the second video camera shooting calculate 3 d space coordinate and the monitoring of invader
The 3 d space coordinate of protected object in region determines the space length of invader and protected object and the movement speed of invader
Degree;The threat level etc. of invader is determined according to the movement velocity of the space length and invader of invader and protected object.
Although the technical program discloses a kind of method of binocular vision detection invader, but it can only have been detected in video image
Invader exists, and in this complex environment of urban highway traffic, the program can only detect road ahead there are barrier,
And detection limit for height barrier in vehicle travel process can not be applied to.
Summary of the invention
The purpose of the present invention is to provide a kind of limit for height method for early warning based on binocular stereo vision, can be in vehicle driving
Limit for height barrier in front of examinations in the process.
For achieving the above object, the present invention provides a kind of limit for height method for early warning based on binocular stereo vision, comprising:
S1. the original image based on binocular camera intake obtains the disparity map of the original image, and sits in the world
The current driving path of the vehicle is obtained in mark system;
S2. the first coordinate that each of disparity map pixel projects in the world coordinate system is calculated;
S3. judge that whether corresponding pixel is in the driving path in the disparity map according to first coordinate
In, if so, the pixel is projected to parallel with the binocular camera image planes in the world coordinate system erect
On straight coordinate plane;
S4. barrier is judged whether there is based on preset height region on the vertical coordinate plane, and if it exists, then obtain
Take the average height of the barrier;
S5. judgement and tracking that step S1 to S4 carries out multiframe to the barrier are repeated, further confirms that the barrier
Hinder object to be limited high to indicate and issue early warning.
According to an aspect of the present invention, it in step S5, repeats step S1 to S4 and barrier progress multiframe is sentenced
Fixed and tracking further confirms that the barrier was limited in high the step of indicating and issuing early warning, judges that the barrier is
No is stationary obstruction, the high mark and by the vehicle body highest of its average height and the vehicle if so, the barrier is limited
Degree of contrast issues limit for height alarm if the average height is lower than the vehicle body maximum height
According to an aspect of the present invention, the world coordinate system is established based on the binocular camera, wherein the generation
The coordinate origin of boundary's coordinate system is vertical point of the optical center of the single camera in the binocular camera to road surface, Y-axis one
Using coordinate origin as starting point, vertically the reference axis with road surface, X-axis are one and are located at road surface simultaneously by starting point of coordinate origin item
And the reference axis vertical with the vehicle central axes, Z axis are one and are located at road surface by starting point of coordinate origin and perpendicular to X
The reference axis of axis.
According to an aspect of the present invention, in step S4, in the world coordinate system, unit depth is obtained from the near to the remote
Locate occupy-place number of the pixel on the vertical coordinate plane in preset height region, and by the occupy-place number and default threshold
Value compares, if the occupy-place number is greater than the preset threshold, there are barriers in present image.
According to an aspect of the present invention, in step S5, comprising:
S51. step S1 to S4 is repeated, if the height region always exists the barrier, obtains the barrier
The first movement velocity, and the second movement velocity of the vehicle according to the CAN bus acquisition of information;
S52. first movement velocity and second movement velocity are compared, if first movement velocity and described
The difference of two movement velocitys is less than preset threshold, then the barrier is stationary obstruction.
According to an aspect of the present invention, first coordinate is calculated using three-dimensional Gaussian meter sciagraphy in step S2.
According to an aspect of the present invention, it in step S2, calculates each of disparity map pixel and projects to institute
In the step of stating the first coordinate in world coordinate system, first coordinate meets:
Wherein, Xt is the coordinate value of the first coordinate X-axis in the world coordinate system, and Yt is that first coordinate exists
The coordinate value of Y-axis in the world coordinate system, Zt are the coordinate value of the first coordinate Z axis in the world coordinate system, xt
It is abscissa value of the pixel in the disparity map, yt is ordinate value of the pixel in the disparity map,
CameraH is the height of the binocular camera, and B is the parallax range of the binocular camera, and f is the binocular camera
Focal length, (u0,v0) be single camera in the binocular camera optical center coordinate, dtIt is the image coordinate in the disparity map
It is the parallax value at midpoint.
According to an aspect of the present invention, in step S2, further includes: obtained in the disparity map according to first coordinate
Each pixel project to the probability distribution value in the world coordinate system, the probability distribution value meets:
Wherein, u is the coordinates measurements of pixel X-axis in the world coordinate system, and v is that pixel is sat in the world
The coordinates measurements of Y-axis in mark system, d are the parallax measurement value of pixel Z axis in the world coordinate system, u0Exist for pixel
The coordinate estimated value of X-axis, v in the world coordinate system0For the coordinate estimated value of pixel Y-axis in the world coordinate system, d0
For the disparity estimation value of pixel Z axis in the world coordinate system.
According to an aspect of the present invention, in step S1, the current traveling road of the vehicle is obtained in world coordinate system
In the step of diameter, the calibration information of CAN bus information and the binocular camera based on vehicle is obtained in world coordinate system
The current driving path of the vehicle.
According to an aspect of the present invention, in step S1, the current traveling road of the vehicle is obtained in world coordinate system
In the step of diameter, using the Z axis in the world coordinate system as symmetry axis, the travelable area with one vehicle width of the vehicle is obtained
Domain, and the driving path is obtained based on the travelable region.
A kind of scheme according to the present invention, through the image projection that acquires binocular camera into world coordinate system, and
The driving path of current vehicle is obtained according to world coordinate system, to ensure that the coordinate of the pixel after projection and the row of vehicle
It sails path to be under same world coordinate system, to ensure that limit for height method for early warning of the invention being capable of accurate judgement source
In the pixel on vehicle running path.By judge some pixel in image coordinate system world coordinates whether there is in
In the driving path of vehicle, it is possible to reduce the pixel number of calculating, to accelerate calculating speed.Particularly with improving, vehicle is real-time
The judgement speed of barrier is advantageous in driving process.In addition, the pixel on the vehicle running path is projected to and binocular
In the parallel numerical coordinates plane (i.e. X/Y plane) of camera image planes, and then it can be convenient and quickly judge in different depths
The height of the pixel of position is spent, and then can be convenient and quickly calculate in present image that whether there are obstacles, is had
Effect simplifies calculating step of the invention, accelerates computational efficiency of the invention, effectively meets in vehicle travel process
Requirement of real-time.
A kind of scheme according to the present invention is detected by the way that the barrier calculated on multiple image is repeated several times by judgement
Whether barrier out is stationary obstruction, so as to effectively reject the interference of moving obstacle etc., is improved of the invention
Detection accuracy ensure that the accuracy of real-time detection of the present invention.When the barrier detected be stationary obstruction and its height it is low
When vehicle body maximum height, early warning can be issued, driver has not only effectively been prompted, ensure that the safety in driving procedure, and
And detecting distance is remote, avoids the case where vehicle and obstacle distance can not closely be braked in time excessively.
Detailed description of the invention
Fig. 1 schematically shows a kind of step block diagram of the limit for height method for early warning of embodiment according to the present invention;
Fig. 2 schematically shows a kind of world coordinate system figure of the limit for height method for early warning of embodiment according to the present invention;
Fig. 3 schematically shows a kind of limit for height method for early warning flow chart of embodiment according to the present invention;
Fig. 4 schematically shows that a kind of binocular camera of embodiment according to the present invention obtains driving path in image
Figure;
Fig. 5 schematically shows that a kind of binocular camera of embodiment according to the present invention obtains the disparity map of image;
Fig. 6 schematically shows a kind of projection of the barrier on driving path of embodiment according to the present invention
Figure.
Specific embodiment
It, below will be to embodiment in order to illustrate more clearly of embodiment of the present invention or technical solution in the prior art
Needed in attached drawing be briefly described.It should be evident that the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skills without creative efforts, can also be according to these
Attached drawing obtains other attached drawings.
When being described for embodiments of the present invention, term " longitudinal direction ", " transverse direction ", "upper", "lower", " preceding ",
" rear ", "left", "right", "vertical", "horizontal", "top", "bottom" "inner", orientation or positional relationship expressed by "outside" are based on phase
Orientation or positional relationship shown in the drawings is closed, is merely for convenience of description of the present invention and simplification of the description, rather than instruction or dark
Show that signified device or element must have a particular orientation, be constructed and operated in a specific orientation, therefore above-mentioned term cannot
It is interpreted as limitation of the present invention.
The present invention is described in detail with reference to the accompanying drawings and detailed description, embodiment cannot go to live in the household of one's in-laws on getting married one by one herein
It states, but therefore embodiments of the present invention are not defined in following implementation.
As shown in Figure 1, a kind of embodiment according to the present invention, a kind of limit for height based on binocular stereo vision of the invention
Method for early warning, comprising:
S1. the original image based on binocular camera intake obtains the disparity map of original image, and in world coordinate system
It is middle to obtain the current driving path of vehicle;
S2. it calculates each of disparity map pixel and projects to the first coordinate in world coordinate system;
S3. judge that corresponding pixel is whether in driving path in disparity map according to the first coordinate, if so, should
Pixel projects on vertical coordinate plane parallel with binocular camera image planes in world coordinate system;
S4. barrier is judged whether there is based on preset height region on vertical coordinate plane, and if it exists, then obtain barrier
Hinder the average height of object;
S5. judgement and tracking that step S1 to S4 carries out multiframe to barrier are repeated, further confirms that the barrier is limited
Height indicates and issues early warning.
As shown in Fig. 2, a kind of embodiment according to the present invention, world coordinate system is established based on binocular camera.At this
In embodiment, to the vertical point on road surface, Y-axis is the coordinate origin O of world coordinate system for midpoint between binocular camera
One using coordinate origin as starting point, vertically the reference axis with road surface, X-axis are one and are located at road surface by starting point of coordinate origin
And the reference axis vertical with vehicle central axes, Z axis are one and are located at road surface by starting point of coordinate origin and perpendicular to X-axis
Reference axis.In the present embodiment, binocular camera is disposed side by side on the middle position of front of the car.By above-mentioned setting,
World coordinate system is established based on binocular camera of the invention, to ensure that limit for height method for early warning of the invention there can be system
One coordinate reference ensure that the image of number camera acquisition carries out the consistency in projection process, ensure that of the invention
The accuracy of limit for height method for early warning.
In conjunction with shown in Fig. 1 and Fig. 3, a kind of embodiment according to the present invention in step S1, is absorbed based on binocular camera
Original image obtain original image disparity map the step of in, obtained respectively by binocular camera two it is distortionless original
Image (i.e. left camera and right camera obtain the two distortionless original images in left and right respectively), and calculated using Stereo matching
Method finds out disparity map (shown in Figure 5) to two distortionless original images of acquisition.In the present embodiment, by using
BoxFilter algorithm obtains disparity map, and the disparity map precision obtained is high and calculating speed is fast.
In conjunction with shown in Fig. 2, Fig. 3 and Fig. 5, a kind of embodiment according to the present invention, in step S1, in world coordinate system
In the step of obtaining the vehicle current driving path, the calibration of CAN bus information and binocular camera based on vehicle is believed
Breath obtains the current driving path of vehicle in world coordinate system.It in the present embodiment, is pair with the Z axis in world coordinate system
Claim axis, obtain have one vehicle width of vehicle travelable region, can running region using the Z axis of world coordinate system as symmetry axis pair
Claim setting.Behind the travelable region for obtaining vehicle, the driving path of vehicle is obtained based on travelable region.In present embodiment
In, the CAN bus based on vehicle effectively supports the serial communications network of distributed AC servo system or real-time control, passes through relevant string
Row communications protocol can obtain the CAN bus information of the i.e. vehicle of relevant information of vehicle in real time, such as: vehicle steering angle,
The information such as throttle, brake.
In conjunction with shown in Fig. 2 and Fig. 4, a kind of embodiment according to the present invention in step S2, is projected using three-dimensional Gaussian meter
Method calculates each of disparity map pixel and projects to the first coordinate in world coordinate system.In the present embodiment, first
Coordinate meets:
Wherein, XtIt is the coordinate value of the first coordinate X-axis in world coordinate system, YtIt is the first coordinate Y in world coordinate system
The coordinate value of axis, ZtIt is the coordinate value of the first coordinate Z axis in world coordinate system, xtIt is abscissa of the pixel in disparity map
Value, ytIt is ordinate value of the pixel in disparity map, cameraH is the height of binocular camera, and B is the base of binocular camera
Linear distance, f are the focal length of binocular camera, (u0,v0) be binocular camera middle position coordinate, dtIt is the figure in disparity map
As the parallax value at coordinate system midpoint.By the measurement error that above reduce the present invention in calculating process, this is improved
The precision of invention.
A kind of embodiment according to the present invention, in step S2, further includes: obtained according to the first coordinate every in disparity map
One pixel projects to the probability distribution value in world coordinate system, and probability distribution value meets:
Wherein, u is the coordinates measurements of pixel X-axis in world coordinate system, and v is pixel Y in world coordinate system
The coordinates measurements of axis, d are the parallax measurement value of pixel Z axis in world coordinate system, u0It is pixel in world coordinate system
The coordinate estimated value of middle X-axis, v0For the coordinate estimated value of pixel Y-axis in world coordinate system, d0It is pixel in world coordinates
The disparity estimation value of Z axis in system.It can be seen that from above-mentioned formula, when the coordinate estimated value of pixel and the measurement of coordinates of pixel
When being worth identical, probability distribution value of the pixel in world coordinate system is maximum.It is shown in Figure 2, the world coordinate system in Fig. 2
In, it is calculated by projecting to the probability distribution in world coordinate system to pixel, obtains its probability distribution curve, be located at curve top
The probability distribution value of end position is maximum, and the probability distribution value positioned at curve bottom position is minimum.
As shown in figure 4, a kind of embodiment according to the present invention, in step S3, by obtaining vehicle CAN bus letter in real time
Breath, obtains the steering wheel angle and speed of vehicle, the width of vehicle is obtained by measurement, thus it is estimated that vehicle
The coordinate of driving path in world coordinate system.If vehicle is kept straight on, driving path is exactly rectangle, if vehicle turning, traveling
Path is exactly concentric circles arc.It is compared and can sentence with pixel coordinate by the coordinate of the driving path of the vehicle estimated
Break and, whether current pixel point is in the driving path of vehicle.
In conjunction with shown in Fig. 2 and Fig. 3, a kind of embodiment according to the present invention, in step S4, in world coordinate system, by
Closely obtain occupy-place number of the pixel on vertical coordinate plane in preset height region at unit depth to remote, and by occupy-place number with
Preset threshold compares, if occupy-place number is greater than preset threshold, there are barriers in present image.In the present embodiment, alive
In boundary's coordinate system, along Z-direction, by closely obtaining unit depth pixel preset height region on vertical coordinate plane to remote
Probability distribution value in (such as range apart from ground 2m to 5m), and the probability distribution value that will acquire be added and obtains occupy-place
Number.By the way that occupy-place number to be compared with preset threshold, if occupy-place number is greater than preset threshold, there are obstacles in present image
Object.When there are barrier, the average height of the barrier is obtained according to coordinate value of the pixel on vertical coordinate plane.Root
According in the disparity map shown in Fig. 5, obtained according to step S1, the probability of the more high then barrier of brightness is bigger, and passes through S3 pairs of step
Whether pixel is judged in driving path in Fig. 5, to generate projected image shown in fig. 6.It is shown in Figure 6,
In present embodiment, if barrier or limit for height mark project on vertical coordinate plane, one will form similar to rodlike
Bright block, the depth of brightness is exactly to represent the height of occupy-place number, rodlike brighter in general, barrier is longer.And common limit for height
In the range of the height of mark is 2m to 5m, therefore use the setting in above-mentioned preset height region can be in order to take ROI to accelerate
Calculating speed.
Through above-mentioned setting, through the image projection that acquires binocular camera into world coordinate system, and according to the world
Coordinate system obtains the driving path of current vehicle, to ensure that the coordinate of the pixel after projection and the driving path energy of vehicle
Enough under same world coordinate system, to ensure that limit for height method for early warning of the invention can be accurately judged to be in vehicle row
The pixel on path is sailed, to reduce calculation amount of the invention.Calculating speed of the invention is improved, particularly with raising
The judgement speed of barrier is advantageous during real-time vehicle running.In addition, the pixel on vehicle running path is projected
Onto the numerical coordinates plane (i.e. X/Y plane) parallel with binocular camera image planes, and then it can be convenient and quickly judge to locate
In the height of the pixel of different depth position, and then it can be convenient and quickly calculate in present image with the presence or absence of barrier
Hinder object, is effectively simplified calculating step of the invention, accelerates computational efficiency of the invention, effectively meet in vehicle driving mistake
Requirement of real-time in journey.
A kind of embodiment according to the present invention, in step S5, comprising:
S51. step S1 to S4 is repeated, if height region always exists barrier, obtains the first movement speed of barrier
Degree, and according to the second movement velocity of CAN bus acquisition of information vehicle.In the present embodiment, step S1 to S4 is repeated, it is right
Multiframe consecutive image is handled, and judges whether preset height region always exists barrier on vertical coordinate plane, if
In the presence of then according to the first movement velocity of multiple image acquisition barrier.Meanwhile according to the of CAN bus acquisition of information vehicle
Two movement velocitys.
S52. the first movement velocity and the second movement velocity are compared, if the difference of the first movement velocity and the second movement velocity
Less than preset threshold, then barrier is stationary obstruction.When barrier is stationary obstruction, according to what is obtained in abovementioned steps
The average height of barrier and the vehicle body maximum height of vehicle compare, if average height is lower than vehicle body maximum height, issue limit
High alarm.In the present embodiment, limit for height alarm can be reported by one of voice, image, text or a variety of combinations
It is alert.
By above-mentioned setting, by the way that the barrier calculated on multiple image is repeated several times, pass through the obstacle for judging to detect
Whether object is stationary obstruction, so as to effectively reject the interference of moving obstacle etc., improves detection accuracy of the invention,
It ensure that the accuracy of real-time detection of the present invention.When the barrier detected is stationary obstruction and its height is lower than vehicle body highest
When height, early warning can be issued, driver has not only effectively been prompted, ensure that the safety in driving procedure, and detecting distance
Far, the case where vehicle and obstacle distance can not closely be braked in time excessively is avoided.
Above content is only the example of concrete scheme of the invention, for the equipment and structure of wherein not detailed description, is answered
When being interpreted as that the existing common apparatus in this field and universal method is taken to be practiced.
The foregoing is merely a schemes of the invention, are not intended to restrict the invention, for the technology of this field
For personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of limit for height method for early warning based on binocular stereo vision, comprising:
S1. the original image based on binocular camera intake obtains the disparity map of the original image, and in world coordinate system
It is middle to obtain the current driving path of the vehicle;
S2. the first coordinate that each of disparity map pixel projects in the world coordinate system is calculated;
S3. judge that corresponding pixel is whether in the driving path in the disparity map according to first coordinate, if
It is that the pixel is then projected to vertical coordinate parallel with the binocular camera image planes in the world coordinate system
In plane;
S4. barrier is judged whether there is based on preset height region on the vertical coordinate plane, and if it exists, then obtain institute
State the average height of barrier;
S5. judgement and tracking that step S1 to S4 carries out multiframe to the barrier are repeated, further confirms that the barrier
The height that is limited indicates and issues early warning.
2. high method for early warning according to claim 1, which is characterized in that in step S5, repeat step S1 to S4 to described
Barrier carries out the judgement and tracking of multiframe, further confirms that the barrier is limited high the step of indicating and issuing early warning
In, judge whether the barrier is stationary obstruction, if so, the barrier is limited high mark and by its average height with
The vehicle body maximum height of the vehicle compares, if the average height is lower than the vehicle body maximum height, issues limit for height alarm.
3. limit for height method for early warning according to claim 1 or 2, which is characterized in that the world coordinate system is based on described double
Mesh camera is established, wherein the coordinate origin of the world coordinate system is the single camera in the binocular camera
Optical center is to the vertical point on road surface, and Y-axis is one, and by starting point of coordinate origin, vertically the reference axis with road surface, X-axis are one to sit
Mark system origin is that starting point is located at road surface and the reference axis vertical with the vehicle central axes, and Z axis is one with coordinate origin
It is located at road surface for starting point and perpendicular to the reference axis of X-axis.
4. limit for height method for early warning according to claim 3, which is characterized in that in step S4, in the world coordinate system,
Occupy-place number of the pixel on the vertical coordinate plane in preset height region at unit depth is obtained from the near to the remote, and
By the occupy-place number compared with preset threshold, if the occupy-place number is greater than the preset threshold, there are obstacles in present image
Object.
5. limit for height method for early warning according to claim 4, which is characterized in that in step S5, comprising:
S51. step S1 to S4 is repeated, if the height region always exists the barrier, obtains the of the barrier
One movement velocity, and the second movement velocity of the vehicle according to the CAN bus acquisition of information;
S52. first movement velocity and second movement velocity are compared, if first movement velocity and second fortune
The difference of dynamic speed is less than preset threshold, then the barrier is stationary obstruction.
6. limit for height method for early warning according to claim 4, which is characterized in that use three-dimensional Gaussian meter sciagraphy in step S2
Calculate first coordinate.
7. limit for height method for early warning according to claim 6, which is characterized in that in step S2, calculate in the disparity map
Each pixel projected in the step of the first coordinate in the world coordinate system, and first coordinate meets:
Wherein, Xt is the coordinate value of the first coordinate X-axis in the world coordinate system, and Yt is first coordinate described
The coordinate value of Y-axis in world coordinate system, Zt are the coordinate values of the first coordinate Z axis in the world coordinate system, and xt is institute
Abscissa value of the pixel in the disparity map is stated, yt is ordinate value of the pixel in the disparity map,
CameraH is the height of the binocular camera, and B is the parallax range of the binocular camera, and f is the binocular camera
Focal length, (u0,v0) be the binocular camera middle position coordinate, dtIt is the image coordinate system midpoint in the disparity map
Parallax value.
8. limit for height method for early warning according to claim 7, which is characterized in that in step S2, further includes: according to described first
Coordinate obtains the probability distribution value that each of disparity map pixel projects in the world coordinate system, the probability
Distribution Value meets:
Wherein, u is the coordinates measurements of pixel X-axis in the world coordinate system, and v is pixel in the world coordinate system
The coordinates measurements of middle Y-axis, d are the parallax measurement value of pixel Z axis in the world coordinate system, u0It is pixel described
The coordinate estimated value of X-axis, v in world coordinate system0For the coordinate estimated value of pixel Y-axis in the world coordinate system, d0For picture
The disparity estimation value of vegetarian refreshments Z axis in the world coordinate system.
9. limit for height method for early warning according to claim 4, which is characterized in that in step S1, obtained in world coordinate system
In the step of vehicle current driving path, the calibration of CAN bus information and the binocular camera based on vehicle is believed
Breath obtains the current driving path of the vehicle described in world coordinate system.
10. limit for height method for early warning according to claim 9, which is characterized in that in step S1, obtained in world coordinate system
In the step of vehicle current driving path, using the Z axis in the world coordinate system as symmetry axis, obtaining has the vehicle
The travelable region of vehicle width, and the driving path is obtained based on the travelable region.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811316705.9A CN109409311A (en) | 2018-11-07 | 2018-11-07 | A kind of limit for height method for early warning based on binocular stereo vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811316705.9A CN109409311A (en) | 2018-11-07 | 2018-11-07 | A kind of limit for height method for early warning based on binocular stereo vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109409311A true CN109409311A (en) | 2019-03-01 |
Family
ID=65472052
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811316705.9A Pending CN109409311A (en) | 2018-11-07 | 2018-11-07 | A kind of limit for height method for early warning based on binocular stereo vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109409311A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109901194A (en) * | 2019-03-18 | 2019-06-18 | 爱驰汽车有限公司 | Onboard system, method, equipment and the storage medium of anticollision |
CN110069990A (en) * | 2019-03-18 | 2019-07-30 | 北京中科慧眼科技有限公司 | A kind of height-limiting bar detection method, device and automated driving system |
CN110401937A (en) * | 2019-04-19 | 2019-11-01 | 王君 | Instant big data transmit-receive platform |
CN110986890A (en) * | 2019-11-26 | 2020-04-10 | 北京经纬恒润科技有限公司 | Height detection method and device |
CN111339956A (en) * | 2020-02-28 | 2020-06-26 | 北京中科慧眼科技有限公司 | Height-limiting early warning method, device and system based on binocular camera and storage medium |
CN111539279A (en) * | 2020-04-14 | 2020-08-14 | 杭州鸿泉物联网技术股份有限公司 | Road height limit height detection method, device, equipment and storage medium |
CN112016366A (en) * | 2019-05-31 | 2020-12-01 | 北京车和家信息技术有限公司 | Obstacle positioning method and device |
CN112364693A (en) * | 2020-10-12 | 2021-02-12 | 星火科技技术(深圳)有限责任公司 | Barrier identification method, device and equipment based on binocular vision and storage medium |
CN112991401A (en) * | 2021-01-20 | 2021-06-18 | 深兰科技(上海)有限公司 | Vehicle running track tracking method and device, electronic equipment and storage medium |
CN113592888A (en) * | 2021-07-01 | 2021-11-02 | 湖北文理学院 | Height limiting frame early warning method, device, equipment and storage medium |
CN114463388A (en) * | 2022-04-13 | 2022-05-10 | 北京中科慧眼科技有限公司 | Binocular camera-based height limiting device detection method and system and intelligent terminal |
CN115100621A (en) * | 2022-08-25 | 2022-09-23 | 北京中科慧眼科技有限公司 | Ground scene detection method and system based on deep learning network |
CN112364693B (en) * | 2020-10-12 | 2024-04-16 | 星火科技技术(深圳)有限责任公司 | Binocular vision-based obstacle recognition method, device, equipment and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120314031A1 (en) * | 2011-06-07 | 2012-12-13 | Microsoft Corporation | Invariant features for computer vision |
CN105825711A (en) * | 2016-02-01 | 2016-08-03 | 乐卡汽车智能科技(北京)有限公司 | Vehicle obstacle early warning method and device |
US20160292519A1 (en) * | 2012-09-19 | 2016-10-06 | Google Inc. | Verifying A Target Object With Reverse-Parallax Analysis |
CN106043121A (en) * | 2016-07-21 | 2016-10-26 | 乐视控股(北京)有限公司 | Height detection early warning system and method and vehicle |
CN108470469A (en) * | 2018-03-12 | 2018-08-31 | 海信集团有限公司 | Road barrier method for early warning, device and terminal |
CN108491810A (en) * | 2018-03-28 | 2018-09-04 | 武汉大学 | Vehicle limit for height method and system based on background modeling and binocular vision |
-
2018
- 2018-11-07 CN CN201811316705.9A patent/CN109409311A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120314031A1 (en) * | 2011-06-07 | 2012-12-13 | Microsoft Corporation | Invariant features for computer vision |
US20160292519A1 (en) * | 2012-09-19 | 2016-10-06 | Google Inc. | Verifying A Target Object With Reverse-Parallax Analysis |
CN105825711A (en) * | 2016-02-01 | 2016-08-03 | 乐卡汽车智能科技(北京)有限公司 | Vehicle obstacle early warning method and device |
CN106043121A (en) * | 2016-07-21 | 2016-10-26 | 乐视控股(北京)有限公司 | Height detection early warning system and method and vehicle |
CN108470469A (en) * | 2018-03-12 | 2018-08-31 | 海信集团有限公司 | Road barrier method for early warning, device and terminal |
CN108491810A (en) * | 2018-03-28 | 2018-09-04 | 武汉大学 | Vehicle limit for height method and system based on background modeling and binocular vision |
Non-Patent Citations (3)
Title |
---|
HERNA´N BADINO 等: "Free Space Computation Using Stochastic Occupancy Grids and Dynamic Programming", 《COMPUTER SCIENCE》 * |
TONG JIA 等: "3D Temperature Distribution Model Based on Vision Method", 《IEEE》 * |
李鹏 等: "一种车载限高交通设施高度自测系统", 《汽车实用技术》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110069990A (en) * | 2019-03-18 | 2019-07-30 | 北京中科慧眼科技有限公司 | A kind of height-limiting bar detection method, device and automated driving system |
CN109901194A (en) * | 2019-03-18 | 2019-06-18 | 爱驰汽车有限公司 | Onboard system, method, equipment and the storage medium of anticollision |
CN110401937A (en) * | 2019-04-19 | 2019-11-01 | 王君 | Instant big data transmit-receive platform |
CN112016366A (en) * | 2019-05-31 | 2020-12-01 | 北京车和家信息技术有限公司 | Obstacle positioning method and device |
CN110986890A (en) * | 2019-11-26 | 2020-04-10 | 北京经纬恒润科技有限公司 | Height detection method and device |
CN110986890B (en) * | 2019-11-26 | 2022-03-25 | 北京经纬恒润科技股份有限公司 | Height detection method and device |
CN111339956A (en) * | 2020-02-28 | 2020-06-26 | 北京中科慧眼科技有限公司 | Height-limiting early warning method, device and system based on binocular camera and storage medium |
CN111539279B (en) * | 2020-04-14 | 2023-08-15 | 杭州鸿泉物联网技术股份有限公司 | Road height limit detection method, device, equipment and storage medium |
CN111539279A (en) * | 2020-04-14 | 2020-08-14 | 杭州鸿泉物联网技术股份有限公司 | Road height limit height detection method, device, equipment and storage medium |
CN112364693A (en) * | 2020-10-12 | 2021-02-12 | 星火科技技术(深圳)有限责任公司 | Barrier identification method, device and equipment based on binocular vision and storage medium |
CN112364693B (en) * | 2020-10-12 | 2024-04-16 | 星火科技技术(深圳)有限责任公司 | Binocular vision-based obstacle recognition method, device, equipment and storage medium |
CN112991401A (en) * | 2021-01-20 | 2021-06-18 | 深兰科技(上海)有限公司 | Vehicle running track tracking method and device, electronic equipment and storage medium |
CN113592888B (en) * | 2021-07-01 | 2024-02-02 | 湖北文理学院 | Height limiting frame early warning method, device, equipment and storage medium |
CN113592888A (en) * | 2021-07-01 | 2021-11-02 | 湖北文理学院 | Height limiting frame early warning method, device, equipment and storage medium |
CN114463388A (en) * | 2022-04-13 | 2022-05-10 | 北京中科慧眼科技有限公司 | Binocular camera-based height limiting device detection method and system and intelligent terminal |
CN115100621A (en) * | 2022-08-25 | 2022-09-23 | 北京中科慧眼科技有限公司 | Ground scene detection method and system based on deep learning network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109409311A (en) | A kind of limit for height method for early warning based on binocular stereo vision | |
Song et al. | Real-time obstacles detection and status classification for collision warning in a vehicle active safety system | |
CN103455144B (en) | Vehicle-mounted man-machine interaction system and method | |
JP3619628B2 (en) | Driving environment recognition device | |
CN110065494B (en) | Vehicle anti-collision method based on wheel detection | |
CN101978404B (en) | Intersection visibility determination device, vehicle with intersection visibility determination device, and method for determining intersection visibility | |
CN104115186B (en) | Stereoscopic article detection device | |
US10713952B2 (en) | Road information recognition system and road information recognition method | |
CN106240458A (en) | A kind of vehicular frontal impact method for early warning based on vehicle-mounted binocular camera | |
CN108037505A (en) | A kind of night front vehicles detection method and system | |
CN105922991A (en) | Lane departure early warning method and system based on generation of virtual lane lines | |
CN108108750A (en) | Metric space method for reconstructing based on deep learning and monocular vision | |
JP2006184276A (en) | All-weather obstacle collision preventing device by visual detection, and method therefor | |
CN110307791A (en) | Vehicle length and speed calculation method based on three-dimensional vehicle bounding box | |
BR112014001837B1 (en) | moving body detection devices and methods | |
CN110456796B (en) | Automatic driving visual blind area detection method and device | |
JP5482672B2 (en) | Moving object detection device | |
WO2014129064A1 (en) | Obstacle detection device and obstacle detection method | |
JP2013061317A (en) | Vehicle surrounding information providing system and method thereof | |
CN108528433A (en) | Vehicle travels autocontrol method and device | |
CN104115188A (en) | Three-dimensional object detection device | |
CN105830131A (en) | Device for detecting the lateral position of a pedestrian relative to the trajectory of the vehicle | |
JP2020504881A (en) | A device operable to determine the position of a part of a lane | |
JP3857698B2 (en) | Driving environment recognition device | |
CN108021899A (en) | Vehicle intelligent front truck anti-collision early warning method based on binocular camera |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190301 |
|
RJ01 | Rejection of invention patent application after publication |