基于视频的交通车流量检测系统的设计,从准确地建立虚拟传感器输出信号与实际车流量的对应关系的角度进行分析与研究,介绍了新的虚拟传感器设计方法,同时提出了信息流的预处理算法和有效信息的判断算法.从而克服了传统设计方法中由于传感器的设计不够合理、信息的获取和判断算法不够科学,导致在有效信息提取过程中消耗大量的软硬件资源等缺点[1].实验结果表明:改进后的设计使整个检测系统对光照变化、道路复杂度以及周围环境有更好的自适应性和鲁棒性.%In the process of video-based traffic detection system design, from the view of establishing the congruent relationship between output signal of virtual sensor and the actual traffic volume, analysis and study is carried out. A new design of the virtual sensor is put forward and pre-processing algorithm for information extraction and judgment method of valid information is improved, so as to avoid some shortcomings of traditional design such as considerable consumption of hardware and software resources in the process of valid information extraction, all of those shortcomings are caused by unreasonable design methods of virtual sensor and unscientific algorithm of information processing in traditional design. The experimental result shows that the revised design has enhanced the adaptability and robustness of the whole system to illumination change, the complexity of the road and the surrounding environment.
展开▼
