期刊名称：International Journal of Future Generation Communication and Networking
印刷版ISSN：2233-7857
出版年度：2016
卷号：9
期号：10
页码：289-298
出版社：SERSC
摘要：With the aggravation of terrorist activities and frequent occuring of all kinds of car bomb attacks, social and public safety issues have become the focus of attention of the whole world. The vehicle chassis foreign body detection system developed in this paper is mainly used in all kinds of important occasions, to achieve real-time and safe detection of vehicles which is helpful to protect the lives and property of citizens, and to block the invasion and transimission of drugs and other harmful substances. Traditional vehicle chassis detection is carried out by security personnel with dedicated portable underbody detection tool, which is time-consuming, laborious and ineffective. In this paper, we construct a intelligent vehicle chassis image detection system. However, there is always distortion in the real time vehicle chassis image captured by multi-line CCD array which therefore needs to be corrected. This paper presents an image correction method for multi linear CCD. Firstly, Sobel differential operator is used to detect the binarized image of the edge of the vehicle chassis on horizontal and vertical directions. Secondly, we use Radon transform to detect angle of inclination of traveling distorted image captured by line scan cameras based on the results of edge detection, using shear transformation to correct binarized chassis image. Finally, a standard chassis image can be obtained based on image interpolation, the characteristic of Sobel operator and recapture the image. Experimental results show that this proposed method is simple and insensitive of stains and light. In addition, a standard chassis image can be obtained after completing the distortion correction.
