文档介绍：上海交通大学
硕士学位论文
D图像的等待车辆数传感器的研究——等待车辆图像的腐
蚀-膨胀识别算法
姓名:陈武
申请学位级别:硕士
专业:机械设计及理论
指导教师:王安麟;马立新
20040201
上海交通大学工学硕士论文
D 图像的等待车辆数传感器的研究
等待车辆图像的腐蚀膨胀识别算法

摘要

道路交通对于一个城市犹如血管神经对于人体一样它的畅通与否直接影响
到整个城市的各项发展进程而城市交通控制系统是一复杂系统具有较强的随
机性突发性和不确定性等特征为了实现城市汽车流实时响应控制建立一个
能较好地反映实际城市交通系统的模型显得尤为重要本课题组提出了基于元胞
自动机(Cellular Automata)方法的城市交通自组织控制模型
本文是基于元胞自动机方法的城市交通自组织控制模型研究的分支为系统
D 待车数传感器的研究通过实时路况图像获得交通路口的
等待车辆数将此信息交由控制系统进行自组织的实时控制规划期间将通过有
效的图像处理去噪均衡化的图像预处理图像增强背景去除二值化处理以
及一定的腐蚀和膨胀效果使系统能检测出实时交通路口的待车数情况针对现有
车辆检测器需对道路进行破坏性改造等问题研究一种基于单幅静态图像处理
D 路口等待车辆数传感器提出在对图像进行有效的预处理的基础上采
用腐蚀膨胀识别算法使系统实时检测出交通路口的等待车辆数并以此作为
城市交通信号控制系统实时决策的状态信息文中通过示例证明了算法的实时
性健壮性及自动化的应用前景

关键词交通工程图像处理腐蚀膨胀算法

5
上海交通大学工学硕士论文
WAITING VEHICLE NUMBER SENSOR
STUDY BASED D IMAGE
---Waiting vehicle image processing by erode-dilate arithmetic

ABSTRACT

Traffic to a metropolis is what blood vessels and nerves to a human body, the
fluency of which will have direct impact on the development of the whole city in
various fields. Due to plexity of the metropolitan control system with its
relatively high randomicity, unexpectedness and uncertainty, it's extraordinarily
important to establish a model that could better reflect the actual traffic in order to
achieve the real time control of the flow of the automobiles in the city. And that is
why the Cellular Automata is proposed as a anizing model for the
metropolitan traffic.
This article works over D sensor for the number of automobiles waiting for
the change of the light at an intersection. It can capture the number of the automobiles
awaiting through a real time picture, and then send the information to the control
system so that the system can process the real time anizing controlling or
planning which includes (the effective treatment of the pictures, the pretreatment of
the pictures to get rid of the noise and to keep equipoise, the stress, the delete of the
b