CN202071799U - Intelligent automobile safety warning device - Google Patents

Abstract

The utility model discloses an intelligent automobile safety warning device which comprises a data collecting module, a data analyzing module, a danger judging module and a warning signal outputting module, wherein, the data collecting module is used for collecting video data through image collecting equipment; the data analyzing module is used for analyzing the collected video data and outputting the relative velocity and the relative distance of a tracking target; and the danger judging module can judge whether the danger of rear-end collision exists or not according to the relative velocity and the relative distance transmitted by the data analyzing module and triggers the warning signal outputting module to output a warning signal if the danger exists. The warning signal can be sent out under the circumstance that an accident probably happens, a corresponding taillight can flicker to warn the driver of a vehicle dangling after, and warning sound can be produced to inform the driver of the own automobile.

Description

The intelligent vehicle safety alarm device

Technical field

The utility model relates to a kind of intelligent vehicle safety alarm device, belongs to the automotive safety technical field.

Background technology

The raising of people's living standard and quality, that automobile becomes is a kind of " accurate popular " consumables.The vehicle of the inside, city is more and more, and city traffic also more and more blocks up.The driving custom of this just " finding time " formula of " robbing runway ", crowded forr a short time just crowded originally running distance.Fearful is, domestic a lot of drivers when driving and the spacing of careless front and back car, particularly metropolitan peak load conditions on and off duty.

When the front vehicle brakes suddenly, closely follow vehicle thereafter owing to there are not enough brake spacings, often lead to the car accident that knocks into the back.According to the U.S.'s statistics in 2007, the vehicle collision rear-end collision that causes because of a variety of causes accounts for about 90% of road traffic accident total amount.In China, just having an appointment 4,000,000,000 yuan in compensated in 2004 according to middle guarantor group 16,000,000,000 yuans is that rear-end collision causes.In November, 2009, Central People's Broadcasting Station points out in reporting rear-end collision: the dead group of the group more than 80% hinders pernicious traffic accident and begins to cause by knocking into the back; Rear-end collision is the arch-criminal who causes city traffic to stop up.

What the safe distance between vehicles of car was generally leaned on before and after judging when the driver drives a vehicle is individual's the experience of driving and sensation.Having done a little investigation recently around, in 18 drivers that the driving experience more than 2 years arranged, when none can ownly certainly be driven a vehicle, is 100% correct to the safety distance judgement of front and back car.That is to say that as long as the fwd car brakes suddenly, they just can not guarantee that the headstock of oneself can " not kiss " front truck " hip "!

The passive type safety method, main to be embodied in automobile collision preventing technical.But the problem that the reality that existing crash-proof technology can not solve well runs into.Most of technology all seriously rely on driver's subjective judgement, such as, by rearview camera is being connected with the read-out of automobile, the driving situation of back is sent on the read-out shows.Like this, the driver in fact to being whether the car of back can knock into the back and knock, neither one is objectively judged.

Lack a kind of dangerous technology of finding: in the process of driving, can carry out real-time supervision and analysis to the situation of this car back, to judge whether to have rear-end impact.

The utility model content

Present technique is a kind of intelligent vehicle safety alarm device that proposes at the problems referred to above, belongs to a kind of active anti-collision technique.

A kind of intelligent safety vehicle alarming device comprises data acquisition module, data analysis module, dangerous judge module, alarm signal output module;

Data acquisition module: be used for gathering video data by image capture device;

Data analysis module: be used for analyzing the relative velocity of output tracking target and relative distance to gathering the video data of coming in;

Dangerous judge module: relative velocity that transmission is come in according to data analysis module and relative distance, judge whether to take place the danger of rear-end impact, if having, then trigger the alarm signal output module, the output alarm signal.

Described intelligent safety vehicle alarming device, described data analysis module comprise that image pretreatment unit, strong unique point processing unit, feature separative element, speed finds the solution unit, feedback unit as a result; Described image pretreatment unit is used for gray processing, the figure image intensifying of image; Strong unique point processing unit adopts the canny operator to carry out rim detection and enhancing; Described feature separative element utilizes pyramid algorith, and strong unique point is followed the tracks of, and separates at the motion feature as the plane according to these unique points, isolates the pursuit-type vehicle that produces relative velocity; Described speed is found the solution the unit, by the unique point of the pursuit-type vehicle of separating, according to the rotation relationship of camera and ground level, obtains the relative velocity of two cars, sends to feedback unit as a result; Described feedback unit as a result, the relative velocity V with processing obtains feeds back to dangerous judge module apart from S.

Described intelligent safety vehicle alarming device, described dangerous judge module are used for relative velocity V and the relative distance S according to input; With predefined threshold values be safety time T _sJudge, if S:V＜T _s, then export energizing signal to the alarm signal output module; Otherwise condition is false, non-output signal.

It can be in the process of driving, in real time dynamic environment (this car back) is carried out video signal collective, by real-time data analysis, the speed of pursuit-type vehicle, distance etc. are carried out comprehensive assessment, to judge whether pursuit-type vehicle the rear-end impact accident can take place in the process of driving.And under the situation that might have an accident, sending alarm signal, the corresponding taillight caution pursuit-type vehicle driver of glimmering sends caution sound prompting Ben Che driver.

Description of drawings

Fig. 1 is the principle schematic of the utility model intelligent vehicle safety alarm device;

Fig. 2 is the utility model data analysis module analysis process figure;

Fig. 3 is the system of 3 axes relation of monocular camera model;

Fig. 4 is the transformation relation of camera coordinates system with world coordinate system;

Fig. 5 is the dangerous judge module decision flow chart of the utility model.

The specific embodiment

Below in conjunction with specific embodiment, the utility model is elaborated.

Embodiment 1

With reference to figure 1, the intelligent vehicle safety alarm device comprises:

Data acquisition module: be used for gathering video data by image capture device (camera).

Data analysis module: be used for analyzing the relative velocity of output tracking target and relative distance to gathering the video data of coming in.

Dangerous judge module: relative velocity that transmission is come in according to data analysis module and relative distance, judge whether to take place the danger of rear-end impact, if having, then trigger the alarm signal output module, the output alarm signal;

The alarm signal output module: when this module is triggered, illustrate that the vehicle of trailing has the danger of rear-end impact, then this module can be sent the corresponding warning signal.

In the running process, gather the environmental information of this car back by vehicle-mounted rearview camera, calculate and analyze the driving situation of pursuit-type vehicle in real time, judge whether pursuit-type vehicle has the danger of rear-end impact.If have, then send the corresponding warning signal.

Embodiment 2

Present embodiment is described in detail at data analysis module.

Because in the process of driving, the vehicle-mounted pick-up head that is installed on this car is constantly to move along with the motion of this car.So the background of the video pictures that transmission is come in also is constantly to move.That is to say that on the angle of the picture of video, background is to move backward.

According to object is different with the motion feature of pursuit-type vehicle in video that produces relative velocity on every side, we can separate both, just can distinguish the vehicle of trailing from video.

The vehicle-mounted pick-up head is demarcated.Through calibrated camera, can set up the contact between " one camera plane, ground ",, can derive as the plane point, in certain time period according to this contact, displacement on ground level (actual road surface) also just can be derived the relative velocity of this vehicle.

Data analysis module comprises that image pretreatment unit, strong unique point processing unit, feature separative element, speed finds the solution unit, feedback unit as a result, and with reference to figure 2, Fig. 2 is data analysis module analysis process figure, may further comprise the steps:

A. the image pretreatment unit is used for gray processing, the figure image intensifying of image, and this mainly is to carry out optimization for further image processing.

B. strong unique point processing unit adopts the canny operator to carry out rim detection and enhancing, and this part work mainly is for next step feature point tracking and separates the unique point that enhancing property is provided.

C. the feature separative element utilizes pyramid algorith, and strong unique point is followed the tracks of, and separates at the motion feature as the plane according to these unique points, isolates the pursuit-type vehicle that produces relative velocity.

D. speed is found the solution the unit, by the unique point of the pursuit-type vehicle of separating, according to the rotation relationship of camera and ground level, obtains the speed of these pursuit-type vehicle unique points at ground level, and then obtains the relative velocity of two cars, sends to feedback unit as a result.Details are as follows for concrete grammar:

Three kinds of system of axess in the monocular camera model concern that as shown in Figure 3 camera coordinates system promptly is to be the system of axes of initial point with optical axis center O, and its z axle satisfies right-hand rule, imaging initial point O _fThe representative plane is photo coordinate system (in the practical application, being coordinate origin with the image upper left corner all), and the actual object system of axes is world coordinate system.

Wherein, P is (X in the value of world coordinate system _w, Y _w, Z _w), P _uBe the subpoint of P at photo coordinate system, the value of its camera coordinates system is (X _u, Y _u, Z _u).θ is that camera coordinates is Z axle and picture plane included angle, and the Z axle is vertical with the picture plane generally speaking, and the θ value is 90 °.And camera coordinates is xOy and picture plane x _fO _fy _fParallel, f is the focal length of camera.

For the conversion that is tied to photo coordinate system from camera coordinates, photo coordinate system is represented with pixel unit, camera coordinates system then is to be that unit represents with the millimeter, therefore, the change process of changing be finish and the pixel unit on picture plane and the linear relationship between the millimeter unit just need be obtained earlier.In Fig. 1, be called principal point with intersection point O ' on the camera optical axis center z direction of principal axis as the plane, coordinate is (c _x, c _y), be pixel unit, and each pixel is at X _fAnd Y _fPhysical dimension be s _x=1/dx and s _y=1/dy, unit are pixels: millimeter, then as the pixel on plane with millimeter between linear relationship suc as formula (1):

$\left(\begin{array}{c}u\\ v\\ 1\end{array}\right)=\left(\begin{array}{ccc}{s}_x& 0& c_x\\ 0& s_y& c_y\\ 0& 0& 1\end{array}\right)\left(\begin{array}{c}x\\ y\\ 1\end{array}\right)---\left(1\right)$

According to projective transformation principle under the little pore model, as the physical coordinates on plane (x, y) cooresponding camera coordinates system satisfies formula (2):

$\left\{\begin{array}{c}x=f\frac{X_u}{Z_u}-f\frac{Y_u}{Z_u\cos \mathrm{\θ}}\\ y=f\frac{Y_u}{Z_u\sin \mathrm{\θ}}\end{array}\right.---\left(2\right)$

Its cooresponding matrix form is formula (3):

$\left(\begin{array}{c}x\\ y\\ 1\end{array}\right)=\frac{1}{Z_u}\left(\begin{array}{cccc}f& -f{\cos }^{-1}\mathrm{\θ}& 0& 0\\ 0& {f\sin }^{-1}\mathrm{\θ}& 0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{c}{X}_u\\ Y_u\\ Z_u\\ 1\end{array}\right)---\left(3\right)$

Simultaneous formula (1) and formula (3) obtain the matrix that formula (4) is camera coordinates system and photo coordinate system conversion.

$\left(\begin{array}{c}u\\ v\\ 1\end{array}\right)=\frac{1}{Z_u}\left(\begin{array}{cccc}{s}_{x}f& -f{\cos }^{-1}\mathrm{\θ}& c_x& 0\\ 0& s_y{f\sin }^{-1}\mathrm{\θ}& c_x& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{c}{X}_u\\ Y_u\\ Z_u\\ 1\end{array}\right)---\left(4\right)$

Wherein, (1/S _x, 1/S _y, c _x, c _y, f θ) is 6 intrinsic parameters of camera, and the matrix of its composition is the intrinsic parameter matrix.

For the conversion that is tied to world coordinate system from camera coordinates, finish by rotation matrix R and translation matrix T, as shown in Figure 4.Wherein, translation matrix T is three dimensional vectors, rotation matrix R be coordinate axle successively around x, y and z axle anglec of rotation ψ,

With formed three the matrix R of τ _x(ψ), R _zTotal product (τ).Their definition is suc as formula (5):

$R_x\left(\mathrm{\ψ}\right)=\left(\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\ψ}& \sin \mathrm{\ψ}\\ 0& -\sin \mathrm{\ψ}& \cos \mathrm{\ψ}\end{array}\right)$

$R_z\left(\mathrm{\τ}\right)=\left(\begin{array}{ccc}\cos \mathrm{\τ}& \sin \mathrm{\τ}& 0\\ \sin \mathrm{\τ}& \cos \mathrm{\τ}& 0\\ 0& 0& 1\end{array}\right)$

Then the computing formula of matrix R is suc as formula (6):

Therefore, the conversion that is tied to world coordinate system from camera coordinates is suc as formula (7), wherein, and 0 ^TExpression (000), R _{3 * 3}Be rotation matrix, Be translation matrix, this transformation matrix is called outer parameter matrix.

$\left(\begin{array}{c}{X}_u\\ Y_u\\ Z_u\\ 1\end{array}\right)=\left(\begin{array}{cc}{R}_{3\×3}& {\stackrel{\→}{T}}_{3\×1}\\ 0^T& 1\end{array}\right)\left(\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right)---\left(7\right)$

At last, simultaneous formula (4) and formula (7) are tried to achieve the transformation relation between photo coordinate system and the world coordinate system, suc as formula (8):

$\left(\begin{array}{c}u\\ v\\ 1\end{array}\right)=\frac{1}{Z_u}\left(\begin{array}{cccc}{s}_{x}f& -f{\cos }^{-1}\mathrm{\&theta;}& c_x& 0\\ 0& s_y{f\sin }^{-1}\mathrm{\&theta;}& {\mathrm{<figure-callout\; id="0"\; label="c\; y"\; filenames="HSA00000511411200022.png"\; state="\{\{state\}\}">c</figure-callout>}}_y& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{cc}{R}_{3\&times;3}& {\stackrel{\&RightArrow;}{T}}_{3\&times;1}\\ 0^T& 1\end{array}\right)\left(\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right)$

$=\left(\begin{array}{cccc}{m}_{00}& m_{01}& m_{02}& m_{03}\\ m_{10}& m_{11}& m_{12}& m_{13}\\ m_{20}& m_{21}& m_{22}& m_{23}\\ m_{30}& m_{31}& m_{32}& m_{33}\end{array}\right)\left(\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right)M_{3\&times;4}\left(\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right)---\left(8\right)$

Wherein, M _{3 * 4}Be perspective projection matrix, the linear relationship in the representation space between three-dimensional point coordinate and the plane of delineation two-dimensional coordinate, (u v 1) ^TExpression P _uPicture plane homogeneous coordinates value, (X _wY _wZ _w1) ^TThe world coordinate system homogeneous coordinates value of expression P.Graphicinformation that obtains based on above geometrical principle and camera model and the relation between the three-dimensional information can be obtained the M of pick up camera _{3 * 4}Matrix.To forward three-dimensional to from two dimension, then need to add a specific 3D modelling.Because in the reality, the point of (highway plane) all drops on the same plane on the ground level, and we can be X-Y (Z=0) plane with this plane, and sets certain the some P on this plane _Wo(X _w=0, Y _w=0, Z _w=0) is the initial point of this three-dimensional coordinate (this can be set).Then the form of the coordinate of this any point x above system of axes all is P _Wx(X _w=x,, Y _w=y, Z _w=0), in fact this be exactly ground level and a specific world coordinate system definite by demarcation are mapped.

By above this relation, can obtain certain the point coordinate point P in image, it is P at the coordinate of world coordinate system _w(X _w, Y _w, Z _w).Then (t1＜t2) cooresponding coordinate is respectively P at continuous time t1, t2 _W1(X _W1, Y _W1, Z _W1), P _W2(X _W2, Y _W2, Z _W2), then try to achieve Δ t=t ₂-t ₂Interior displacement difference:

ΔS＝P _w1(X _w1，Y _w1，Z _w1)-P _w2(X _w2，Y _w2，Z _w2)

By Δ S and Δ t, can obtain speed V:

$V=\frac{\mathrm{\&Delta;S}}{\mathrm{\&Delta;t}}=[P_{w1}(X_{w1},Y_{w1},Z_{w1})-P_{w2}(X_{w2},Y_{w2},Z_{w2})]/\mathrm{\&Delta;t}$

In fact this V is exactly the relative velocity of two cars.

Because the actual coordinate obtained like this is initial point that above figure sets is reference, if measure this initial point actual distance S apart from camera when setting ₁, then by the relative coordinate P that can obtain a P _w(X _w, Y _w, Z _w) with the actual distance S (being the relative distance of tracking target and this car) of camera:

$S=S_1\&PlusMinus;\sqrt{{X_w}^2+{Y_x}^2+O^2}$

E. feedback unit as a result, the relative velocity V with processing obtains feeds back to dangerous judge module apart from S.

Judge whether to send energizing signal by dangerous judge module and give the alarm signal output module, for example can send optical signal by automobile tail light, prompting back car is noted maintaining safe distance, and can make a kind of automobile tail light with safety warning function like this.With reference to figure 5 are diagram of circuits of dangerous judge module, according to importing data: relative velocity V, relative distance S; With predefined threshold values be safety time T _s(this can set according to actual conditions).If S:V＜T _sThen export energizing signal to the alarm signal output module; Otherwise condition is false, non-output signal.

Should be understood that, for those of ordinary skills, can be improved according to the above description or conversion, and all these improvement and conversion all should belong to the protection domain of the utility model claims.

Claims (2)

1. an intelligent safety vehicle alarming device is characterized in that, comprises data acquisition module, data analysis module, dangerous judge module, alarm signal output module, data acquisition module: be used for gathering video data by image capture device;

Data analysis module: be used for analyzing the relative velocity of output tracking target and mouthful relative distance to gathering the video data of coming in;

Dangerous judge module: relative velocity that transmission is come in according to data analysis module and relative distance, judge whether to take place the danger of rear-end impact, if having, then trigger the alarm signal output module, the output alarm signal.

2. intelligent safety vehicle alarming device according to claim 1 is characterized in that, described data analysis module comprises that image pretreatment unit, strong unique point processing unit, feature separative element, speed finds the solution unit, feedback unit as a result; Described image pretreatment unit is used for gray processing, the figure image intensifying of image; Strong unique point processing unit adopts the canny operator to carry out rim detection and enhancing; Described feature separative element utilizes pyramid algorith, and strong unique point is followed the tracks of, and separates at the motion feature as the plane according to these unique points, isolates the pursuit-type vehicle that produces relative velocity; Described speed is found the solution the unit, by the unique point of the pursuit-type vehicle of separating, according to the rotation relationship of camera and ground level, obtains the relative velocity of two cars, sends to feedback unit as a result; Described feedback unit as a result, the relative velocity V with processing obtains feeds back to dangerous judge module apart from S.

Images