基于嵌入式系统的公交车实时客流量检测系统的设计与实现

发布时间：2018-03-29 06:27

本文选题：实时客流量检测　切入点：运动目标跟踪　出处：《内蒙古大学》2017年硕士论文

【摘要】：智能调度是智能公交系统(ITS)的关键组成部分,对于改善城市交通堵塞、提高公共资源利用率、降低公共交通系统运营成本和方便市民出行等具有极其重要的作用,而实时准确的公交车客流量信息是实现智能调度的最重要的数据依据。本文移植了 Linux 操作系统、OpenCV(Open Source Computer Vision Library)计算机视觉库、Qt图形界面库(跨平台的C++图形用户界面应用程序框架)等到ARM嵌入式平台上,针对不同客流密度时乘客上下车的场合,提出了不同的统计方法,完成了公交车客流量统计系统的设计和实现,最后通过GPRS将统计信息发送到指定了 IP地址的数据库上。本文先详细介绍了系统软硬件环境的搭建,包括Linux内核、CRAMFS根文件系统移植到ARM平台,OpenCV计算机视觉库、QtE图形界面库的移植等,再针对公交车乘客上下车的场景,设计了客流量统计系统。在对客流量进行统计时,本系统针对公交车客流密度大小的不同场合分别进行统计。在客流量密集的场合,乘客之间互相挨在一起,此时的前景检测图像将是一片互相粘连的区域,对运动目标的分割和检测也无法完成。为解决这个问题,本文提出使用霍夫圆变换、质心坐标、颜色直方图作为特征,对乘客进行检测和跟踪的方法,并通过实验来验证此方法。实验结果表明,本文提出的方法解决了在客流量密集的场合下无法对运动目标进行分割的问题,并具有较好的实时性和鲁棒性,对于公交车实时客流量统计的研究做了补充,具有重要的现实意义。
[Abstract]:Intelligent dispatching is a key component of Intelligent Public Transportation system (ITS), which plays an extremely important role in improving urban traffic congestion, increasing the utilization of public resources, reducing the operating cost of public transportation system and making it convenient for citizens to travel. The real time and accurate bus passenger flow information is the most important data basis for intelligent dispatching. In this paper, the computer vision library of Linux operating system, Open Source Computer Vision Library, is transplanted to QT graphical interface library (C graphical user boundary across platforms). On the ARM embedded platform, According to the situation of passengers getting on and off with different passenger flow density, different statistical methods are put forward, and the design and implementation of bus passenger flow statistics system are completed. Finally, the statistical information is sent to the database with specified IP address through GPRS. Firstly, this paper introduces the construction of the software and hardware environment of the system in detail. Including the transplantation of the Linux kernel CRAMFS root file system to the ARM platform computer vision library and QtE graphical interface library, and the design of the passenger flow statistics system for the scene of bus passengers getting on and off. The system makes statistics for different occasions of bus passenger flow density. In the case of dense passenger flow, passengers are close to each other. At this time, the foreground detection image will be an area where each other adheres to each other. In order to solve this problem, this paper proposes a method to detect and track passengers by using Hoff circle transformation, centroid coordinates and color histogram as features. The experimental results show that the proposed method solves the problem that the moving target can not be segmented in the case of dense passenger flow, and has good real-time and robustness. It is of great practical significance to supplement the research of bus real-time passenger flow statistics.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U495;TP391.41

【相似文献】