Machine-vision-based electronic automobile insurance fee meter

Abstract

The present invention provides a machine-vision-based electronic automobile insurance fee meter. The machine-vision-based electronic automobile insurance fee meter can capture a lane departure behavior, and record a distance and duration of a lane departure in a memory module; also can capture a following-too-close behavior, and record a distance and duration of dangerous following in the memory module; and also can calculate a journey-based mileage insurance fee, a total mileage insurance fee P being a product of a mileage L of a journey, a unit mileage (per kilometer) insurance rate R.sub.km and a safety floating factor f. The present technical solution charges the insurance fee according to the actual driving mileage, makes the charging of the insurance fee fairer and more reasonable, and encourages safe driving and driving less; and also can identify the most important dangerous driving risks including following-too-close and lane departure, making the pricing of the insurance fee more accurate and reasonable.

Claims

1. A machine-vision-based electronic automobile insurance fee meter, comprising: a central processing unit, a memory module, a sensor module, a user interface module and a network communication module, the sensor module including: a camera module, an accelerometer/angular velocity meter unit and a GPS module, wherein the camera module consists of a camera unit and an image processing chip; and the meter is configured to acquire image information of the front of an insured car by the camera unit, the image processing chip processes and identifies features of the image information, the central processing unit makes a determination on the identified features, and stores times information and time information which accord with lane departure and following-too-close behaviors in the memory module; wherein the meter is configured such that the central processing unit calculates the safety floating factor f of each journey, $\begin{array}{c}f=f\left(\mathrm{CDN},\mathrm{ACDT},\mathrm{LDN},\mathrm{ALDT}\right)\\ =f_{\mathrm{CDN}}{f}_{\mathrm{ACDT}}{f}_{\mathrm{LDN}}{f}_{\mathrm{ALDT},}\end{array}$ wherein CDN represents 100-kilometer dangerous following times, ACDT represents average dangerous following duration, LDN represents 100-kilometer lane departure times, ALDT represents average lane departure duration, $f_{\mathrm{CDN}}=100\frac{\underset{t}{.Math.}{T}_{\mathrm{KD}}}{2}\mathrm{total}\mathrm{journey}\mathrm{mileage}$ $f_{\mathrm{ACDT}}=\frac{2\underset{t}{.Math.}{T}_{\mathrm{KD}}}{\underset{t}{.Math.}{\mathrm{SK}}_t}$ $f_{\mathrm{LDN}}=100\frac{\underset{t}{.Math.}{\mathrm{SD}}_t}{2}\mathrm{total}\mathrm{journey}\mathrm{mileage}$ $f_{\mathrm{ALDT}}=\frac{2\underset{t}{.Math.}{T}_{\mathrm{DD}}}{\underset{t}{.Math.}{\mathrm{SD}}_t}.$

2. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the accelerometer/angular velocity meter unit acquires an acceleration and an angular velocity of the insured car; and the GPS module acquires data of GPS, GLONASS or Beidou system real-time trajectory of the insured car.

3. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the user interface module includes: a display module and a key unit; the user interface module is configured such that a user can select and query a kilometer insurance fee unit price, a charged mileage, an insurance fee of the charged mileage, an insurance fee floating factor and lane departure and following-too-close records by operating the key unit; and query results are displayed on the display module.

4. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the network communication module is configured to support 4G/3G/2G network systems, and is used for carrying out information interaction with a car insurance business database of an insurance company and generating an insurance fee transaction record.

5. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the meter is configured such that gray scale filtering processing is performed on the image information by the image processing chip; the central processing unit detects out a quadrilateral contour TT B B of a rear portion of a car in the front according to bilateral symmetrical and horizontal-vertical boundary features of the car in the front, and obtains coordinates of four vertexes of the contour in an image, then calculates a contour feature ratio, an upper-lower bottom ratio R.sub.b and a height-width ratio R.sub.l, $R_b=\frac{{\mathrm{TT}}^{}}{{\mathrm{BB}}^{}},R_l=\frac{\mathrm{TB}}{{\mathrm{BB}}^{}},$ and a car feature table is queried according to values of R.sub.b and R.sub.l so as to obtain a type and a width b.sub.F of the car in the front; and finally, according to a focus length and an image resolution of the camera module, an actual distance Z from the car in the front is calculated, $Z=\frac{b_{F}f}{{\mathrm{BB}}^{}/HR},$ wherein b.sub.F represents a car width of the car in the front, f represents a lens focus length of a camera, BB represents a pixel width in the image of the car in the front, and HR represents a horizontal resolution of the camera unit; and if a current actual car velocity is Vt, and according to set minimum safe following reaction time T.sub.s, when Z<V.sub.tT.sub.s, a dangerous following behavior is determined.

6. The machine-vision-based electronic automobile insurance fee meter according to claim 5, characterized in that the meter is further configured such that the central processing unit calculates following-too-close duration T.sub.KD,
T.sub.KDSK.sub.t*dt, wherein SK.sub.t represents a following safety state, ${\mathrm{SK}}_t=\{\begin{array}{cc}1\left(\mathrm{too}\mathrm{close}\right)& \mathrm{when}Z<V_t{T}_s\\ 0\left(\mathrm{safe}\right)& \mathrm{when}Z{V}_t*T_s\end{array},$ and when the following-too-close duration T.sub.KD is greater than a preset value, the dangerous following behavior is determined.

7. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the meter is configured such that gray scale filtering processing is performed on the image information by the image processing chip; and the central processing unit identifies a laneway line sign in a current road and coordinates of the laneway line sign in the image according to edge features of an oblique straight line of the laneway line sign, calculates coordinates of a lane centerline OO, and meanwhile, also calculates a coordinate distance LR of left and right laneway lines and a lane width w, compares coordinate differences between the lane centerline OO in the image and a central axis MM of front and rear wheels of the car, i.e., OM and OM, and calculates to obtain corresponding far-end and near-end central axis departures d2 and d1 through an optical perspective geometric imaging formula, $d2=\frac{O^{}{M}^{}/\mathrm{HR}}{f}Z2,$ wherein OM represents a far-end image coordinate pixel distance value, Z2 represents a distance from a far-end line to the camera, HR represents an image resolution of the camera and f represents a focal length of the camera; $d1=\frac{\mathrm{OM}/\mathrm{HR}}{f}Z1,$ wherein OM represents a near-end image coordinate pixel distance, Z1 represents a distance from a near-end line to the camera, HR represents the image resolution of the camera, and f represents the focal length of the camera; then according to the distances Z2 and Z1 from the far-end and near-end lines to the camera, on the basis of a geometric proportion relationship, a transverse position d of an axle center C of the front wheels of the car in a lane is calculated; $d=\left(d1-d2\right)\frac{Z1}{Z2-Z1}+d1$ and finally, according to the lane width w and a car width b, an actual distance of departure dp of the car from the lane line can be calculated, $\mathrm{dp}=\{\begin{array}{cc}d+\frac{b}{2}-\frac{w}{2},& \mathrm{when}d>\frac{w-b}{2}\\ d-\frac{b}{2}+\frac{w}{2},& \mathrm{when}d<\frac{b-w}{2}\\ 0,& \mathrm{when}\frac{b-w}{2}d\frac{w-b}{2}\end{array},$ wherein the lane width w is $w=\frac{\mathrm{LR}}{\mathrm{OM}},$ and when the lane departure distance dp is greater than a preset value, a lane departure behavior is determined.

8. The machine-vision-based electronic automobile insurance fee meter according to claim 7, characterized in that the meter is further configured such that the central processing unit calculates the lane departure duration T.sub.DD,
T.sub.DD|SD.sub.t*dt| wherein a land hold state SD.sub.t is ${\mathrm{SD}}_t=\{\begin{array}{cc}1\left(\mathrm{right}\mathrm{departure}\right)& \mathrm{when}\mathrm{dp}>0\\ 0\left(\mathrm{safe}\right)& \mathrm{when}\mathrm{dp}=0\\ -1\left(\mathrm{left}\mathrm{departure}\right)& \mathrm{when}\mathrm{dp}<0\end{array}$ and when the land hold state SD.sub.t is greater than a preset value, the lane departure behavior is determined.

9. The machine-vision-based electronic automobile insurance fee meter according to claim 1, characterized in that the meter is configured such that the central processing unit calculates a journey-based mileage insurance fee,
P=R.sub.kmLf wherein a mileage L is $L=\underset{i=1}{\overset{n}{.Math.}}{L}_{i-1,i}$ wherein L.sub.i1,i represents a distance from a trajectory point i1 to a trajectory point i.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a module diagram of a machine-vision-based electronic automobile insurance fee meter according to one preferred embodiment of the present invention;

(2) FIG. 2 is a schematic diagram of a central processing unit determining a following-too-close behavior according to one preferred embodiment of the present invention; and

(3) FIG. 3 is a schematic diagram of a central processing unit determining a lane departure according to one preferred embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(4) As shown in FIG. 1, according to the present invention, a machine-vision-based electronic automobile insurance fee meter includes a central processing unit, a memory module, a sensor module, a user interface module and a network communication module. The sensor module includes: a camera module, an accelerometer/angular velocity meter unit and a GPS module, wherein the camera module consists of a camera unit and an image processing chip. The meter is configured to acquire image information of the front of an insured car by the camera unit. The image processing chip performs identification on features of the image information. The central processing unit makes a determination on the identified features, and store times information and time information which accord with lane departure and following-too-close behaviors in the memory module.

(5) An example is provided below. The machine-vision-based electronic automobile insurance fee meter is mounted at a position of a front windshield at a central axis of an automobile. As shown in FIG. 2, a car width of the car is 1.8 meters, and a far-end line distance Z2 of a trapezoid of a reference lane is set as 10 meters, a near-end line distance Z1 of the trapezoid of the reference lane is set as 5 meters, an image resolution HR of a camera is set as 450 pixel/mm, and a focal length is set as 2.8 mm.

(6) The meter captures a contour of a rear portion of a car in the front, image lengths of TT and B B are 102 pixel, image lengths of TB and TB are 144 pixel, a height-width ratio R.sub.l is equal to 1.41, and an upper-lower ratio R.sub.b is equal to 1.0. A car contour feature table is queried, so that the car in the front can be determined as a passenger car and the car width b.sub.F is 2.5 meters.

(7) TABLE-US-00001 Height-Width Upper-Lower Car Width Car Model Ratio R.sub.i Ratio R.sub.b b.sub.F (meter) Sports Car 0.5-0.7 <0.98 2.0 Sedan 0.7-0.8 1.8 SUV/MPV 0.8-0.95 1.9 Truck 0.95-1.4 0.98 2.5 Passenger Car 1.4-1.7 2.5

(8) A car distance Z is calculated,

(9) $Z=\frac{b_{F}f}{{\mathrm{BB}}^{}/\mathrm{HR}}=\frac{2.5*2.8}{102/450}=27.57\left(\mathrm{meter}\right).$

(10) When a current car velocity of the car is 60 km/s and dangerous following duration is set as 2.5 s, then

(11) a following state SK.sub.t=1(too close) because Z=27.57<V.sub.tT=41.67.

(12) As shown in FIG. 3, through image identification, a width of the reference lane LR is 750 pixel, a near-end central axis OM is shifted by 240 pixel, a far-end central axis O M is shifted by 110 pixel.

(13) A far-end transverse departure distance d2 is

(14) $d2=\frac{O^{}{M}^{}/\mathrm{HR}}{f}Z2=10*\frac{110/450}{2.8}=0.873\left(\mathrm{meter}\right).$

(15) A near-end transverse departure distance d1 is

(16) $d1=\frac{\mathrm{OM}/\mathrm{HR}}{f}Z1=5\frac{240/450}{2.8}=0.952\left(\mathrm{meter}\right).$

(17) A lane line width is

(18) $w=\frac{\mathrm{LR}}{\mathrm{OM}}=2.98\left(\mathrm{meter}\right).$

(19) An automobile axle center transverse departure distance d is

(20) $d=\left(d1-d2\right)\frac{Z1}{Z2-Z1}+d1=1.03\left(\mathrm{meter}\right).$

(21) A distance of departure dp of the car from the lane line is

(22) $\mathrm{dp}=d+\frac{b}{2}-\frac{w}{2}=1.03+0.9-1.49=0.44\left(\mathrm{meter}\right).$

(23) A land hold state is determined as
lane hold state SD.sub.t=1(right departure), because dp>0.

(24) An insurance fee calculation example is provided. A total mileage of a current journey the car is 22.8 kilometers, and a per kilometer insurance rate R.sub.km is set as 0.5 yuan. In the journey, dangerous following times of a driver are 2 times, and average dangerous following duration is 12.5 s; and lane departure times are 12 times, and average lane departure duration is 3.2 s.

(25) Specific following and lane departure indexes of the journey are as follows: CDN=8.8 ACDT=12.5 LDN=52.6 ALDT=3.2

(26) The following factor tables are queried.

(27) TABLE-US-00002 CDN 100-kilometer Following-too-close Times (times) f.sub.CDN Factor 0 0.25 1-2 0.3 3-5 0.56 6-15 0.75 16-36 1 37-64 1.38 >64 1.8

(28) TABLE-US-00003 ACDT Average Following-too-close Duration (s) f.sub.ACDT Factor 0 0.4 (0, 2) 0.6 [2, 4) 0.85 [4, 10) 1 [10, 30) 1.15 [30, 60) 1.4 [60, +) 2

(29) TABLE-US-00004 LDN 100-kilometer Lane Departure Times (times) f.sub.LDN Factor 0 0.4 1-10 0.6 11-20 0.85 30-60 1 60-100 1.15 100-160 1.4 >160 2

(30) TABLE-US-00005 ALDT Average Lane Departure Duration (s) f.sub.ALDT Factor 0 0.7 (0, 2) 0.85 [2, 4) 1 [4, 10) 1.15 [10, 30) 1.4 [30, 60) 1.8 [60, +) 2.5

(31) A safety floating factor f of a driving behavior of the journey is

(32) $\begin{array}{c}f=f\left(8.8,12.5,52.6,3.2\right)\\ =0.751.1511=0.8625.\end{array}$

(33) The total mileage insurance fee P of the journey is
P=R.sub.kmLf=0.522.8*0.8625=9.8323(yuan).

(34) The preferred specific embodiments of the present invention are described above in detail. It should be understood that those skilled in the art could make various modifications and changes according to the concept of the present invention without inventive skills. Therefore, any technical solution that can be obtained by a person skilled in the art by logic analysis, reasoning or limited experiments according to the concept of the present invention and on the basis of the prior art shall fall within the protective scope determined by the appended claims.