CAJ | 학술논문

视频交通信息检测近年发展很快，因具有大区域检测、设置灵活等优越性，已成为智能交通系统的一个研究热点。在以往的检测基础上提出一种基于虚拟线框的背景差分法进行车辆识别、车辆行驶方向判断及速度计算等算法。应用正交实验法来确定虚拟线框的大小，采用选择-平均法实现动态的背景更新。最后给出系统检测结果，经分析得出该系统准确率高于80o。
시빈교통신식검측근년발전흔쾌，인구유대구역검측、설치령활등우월성，이성위지능교통계통적일개연구열점。재이왕적검측기출상제출일충기우허의선광적배경차분법진행차량식별、차량행사방향판단급속도계산등산법。응용정교실험법래학정허의선광적대소，채용선택-평균법실현동태적배경경신。최후급출계통검측결과，경분석득출해계통준학솔고우80o。
The video-based traffic detection has developed quickly,since it has many advantages, for instance, wider-area detection and superior flexibility, it is becoming a pop area in the intelligent transportation system. On the basic of the other methods for video traffic detection, this paper provides a detailed description of distinguishing from different vehicles, background update,direction and speed calculation of the vehicle,using the background subtraction between the background image and the image in the virtual frame. In the paper the orthogonal experimentation has been used to ascertain the size of the frame,and used the selective-average method to realize the dynamic background updating. Finally, the paper gives the results of the system and analyzes to elicit that the system has a high veracity to 80o.