Design And Implementation Of Warning System For Front Vehicle's Status Based On Image

With the increasing number of vehicles on road in China,it puts more and more pressure on road traffic safety.A great quantity of casualties and huge economic losses are caused by rear-end collisions every year.Thus,it's really vital and urgent to improve the domestic traffic safety environment by reducing or avoiding such kind of accidents.This paper proposes a kind of warning system based on image to monitor front vehicle's status,which can assist the drivers to recognize the operations of the front vehicles.The warning system needs to recognize the location of vehicle's license plate and the status of front vehicle's lamps by image processing which is captured by a image sensor and includes the information about front vehicles and road conditions.The main contents of this paper is just as following:First of all,the advantages and the disadvantages of the existing mainstream machine vision method in field of road traffic are given based on the analysis of the applications of machine vision in the field of intelligent transportation.Moreover,the contents and schemes of the thesis are also determined.Secondly,this paper proposes a solution for the warning system based on the analysis of the feasibilities and costs due to the requirements of detecting front vehicle's status,which clarifies the functions of image acquisition unit,image processing unit and warning unit as well as their relationship in the system.Thirdly,for the hardware of the image acquisition and processing unit in the system,this paper also provides the compositions and development environment of the image acquisition system based on the module of processor platform HI3518E and the image sensor OV9712.Then,for the lamps' recognition in the system,a simple license plate detecting algorithm is proposed to identify the front vehicle's license plate area quickly and effectively,so as to obtain the position of front vehicle's lamps.The validity of the method is verified by several experiments.Finally,aiming at the states of lamps in the vehicle's status recognition algorithms,a method is proposed to judge if the lamps are on or off according to the brightness information of the vehicle's images in HSI color space,and the effectiveness of the method is also verified by several experiments.In summary,the warning system of front vehicle's status proposed in this paper can detect the status of front vehicles and inform the drivers in some degree,which can reduce or avoid the occurrence of rear-end accidents effectively.