基于ROI最长直线提取算法的车辆超限检测系统研究

Vehicle Overrun Detection System Research Based on Maximum ROI Line Extraction Algorithm[HS)]

摘要:

针对当前公路上车辆超限而导致交通安全隐患问题,提出了一种基于图像测量的车辆超长、超宽、超高检测系统；利用CCD摄像头捕捉限速车道内行驶的车辆的图片,经过图像分类、背景去除、噪声去除提高图像清晰度,按ROI(region of interest)最长直线提取算法提取ROI内车辆轮廓的最长像素直线,通过相机标定获取车辆实际长宽高信息；系统对于车辆超限能够实现不停车检测,有效提高了公路通行速率；实验数据表明,该系统对于不同类型车辆,均具有良好的准确性和鲁棒性。

Abstract:

In view of the current highway vehicles overrun and cause problem such as traffic safety hidden trouble, a system based on digital image processing that measure the vehicles’ surpass length, extra-wide, ultra-high is proposed. Using CCD camera to capture the slowly moving vehicles image in the driveway, after image classification, background removal, noise removal to improve image clarity, according to the maximum ROI line extraction algorithm to extract ROI longest pixel lines, then through the camera calibration to get vehicle actual dimensions information. The system for the overrun vehicle’s detection without parking, effectively improve the rate of road traffic. The experimental data show that the system for different types of vehicles, has good accuracy and robustness.