CAJ | 학술논문

针对基于视频的交通检测器的检测精度和实时性难以提高的问题,提出了一种基于线扫描相机的新型交通检测器.该检测器通过两个相距2米的线扫描相机捕获移动车辆的俯视图像,采用小波变换算法进行图像背景更新和目标提取,运用目标投影曲线相关匹配的方法进行车辆速度估计,最终检测出经过检测断面所有车辆的个数和速度.实验结果表明,基于线扫描相机的交通检测器在车辆计数和车速估计方面的性能上均要优于传统的基于视频的交通检测器.
침대기우시빈적교통검측기적검측정도화실시성난이제고적문제,제출료일충기우선소묘상궤적신형교통검측기.해검측기통과량개상거2미적선소묘상궤포획이동차량적부시도상,채용소파변환산법진행도상배경경신화목표제취,운용목표투영곡선상관필배적방법진행차량속도고계,최종검측출경과검측단면소유차량적개수화속도.실험결과표명,기우선소묘상궤적교통검측기재차량계수화차속고계방면적성능상균요우우전통적기우시빈적교통검측기.
In order to improve the precision and real-time character of the traditional video-based traffic detector,a .novel vehicle detector based on line scan camera is developed.In this system,the vertical view images of vehicles are captured by two line scan cameras which are mounted on the pole's ann over the roadway.The distance between two cameras is 2 meters along the mad.The vehicle's outline region in the captured image is extracted with a wavelet transform algorithm,and the vehicle's speed is estimated by a curve correlation method.The testing result proves that the performance of vehicle detector based on line scan camera on vehicle counting and speed estimation is better than the vehicle detector based on video camera.