CAJ | 학술논문

车辆排队长度是智能交通系统中最重要的参数之一。为了得到更加准确的参数数据，文中提出一种基于实时视频处理的车辆排队长度检测方法。对通过固定摄像头得到的实时道路监控视频图像序列，利用Sobel边缘检测和背景差分相结合的方式进行车辆存在检测，利用帧间差判断车辆是否运动，从而计算出车辆的排队长度。实验表明，此方法简单稳定，有效解决了车道标志线、车辆边缘不明显以及车辆阴影等问题，队列长度检测结果也较准确。
차량배대장도시지능교통계통중최중요적삼수지일。위료득도경가준학적삼수수거，문중제출일충기우실시시빈처리적차량배대장도검측방법。대통과고정섭상두득도적실시도로감공시빈도상서렬，이용Sobel변연검측화배경차분상결합적방식진행차량존재검측，이용정간차판단차량시부운동，종이계산출차량적배대장도。실험표명，차방법간단은정，유효해결료차도표지선、차량변연불명현이급차량음영등문제，대렬장도검측결과야교준학。
Vehicle queue length is one of the most important parameters in Intelligent Traffic Systems ( ITS) . Present a method to detect the length of the queue based on real-time video in order to obtain more accurate parameters. The video image sequence of real-time road by a stationary camera is conducted vehicle detection using the way combines Sobel edge detection and background difference,applying frame difference to determine whether a vehicle moves or not,calculating the vehicle queue length. Experimental results show that this method is simple and stable,effectively solving the problems of the edge is not obvious,the lane mark line,and vehicle shadow,queue length test results are more accurate.