【摘要】：智能交通系統(tǒng)(ITS)是將先進的科學技術相融合并有效地運用于在交通管理、交通信息服務和車輛控制等方面的綜合交通管理系統(tǒng)。它增強了車輛和道路與使用者之間的聯(lián)系,是一種大規(guī)模、全方位、實時、準確、高效的綜合智能交通管理系統(tǒng),對交通管理和控制有著重要的意義和作用。本論文設計的總體目標是通過圖像處理算法提取車隊的縱深長度或車輛的個數(shù),并將其作為系統(tǒng)控制的參數(shù)來控制調(diào)整交通信號顯示方式。全部的圖像處理算法程序?qū)⒁浦驳揭訟RM為核心的交通控制板上,從而實現(xiàn)一種自適應的節(jié)點智能交通控制方法。本課題主要研究一種簡單有效的交通圖像采集和處理算法,通過圖像處理算法提取車隊排隊信息。在算法上,首先選擇可以突顯車輛并去除大量噪聲的預處理方法,選擇簡單、快捷的背景差分法作為提取車輛的方法。在背景差分法中增添背景更新模型和分割算法,并對分割算法進行了改進,使算法運行效率更高�？紤]到實時性的問題,提出了車道提取方法,從而提高了系統(tǒng)的實時性。對于車隊長度和車輛個數(shù)的統(tǒng)計,建立了行坐標模型,實現(xiàn)了圖像坐標系和實際坐標系的距離轉(zhuǎn)換,從而估算出車隊長度,為后續(xù)硬件系統(tǒng)實現(xiàn)提供了重要參數(shù)
[Abstract]:Intelligent Transportation system (ITS) is an integrated traffic management system, which integrates advanced science and technology and is effectively used in traffic management, traffic information service and vehicle control. It enhances the connection between vehicles, roads and users. It is a large-scale, omni-directional, real-time, accurate and efficient integrated intelligent traffic management system, which is of great significance and effect to traffic management and control. The overall goal of this paper is to extract the longitudinal length or the number of vehicles by image processing algorithm, and use it as the parameters of system control to control and adjust the display mode of traffic signals. All the image processing algorithm programs will be transplanted to the traffic control board with ARM as the core so as to realize an adaptive node intelligent traffic control method. In this paper, a simple and effective traffic image acquisition and processing algorithm is studied, and the queue information is extracted by image processing algorithm. In the algorithm, firstly, the pretreatment method which can highlight the vehicle and remove a lot of noise is selected, and the simple and fast background difference method is chosen as the method to extract the vehicle. The background update model and segmentation algorithm are added to the background difference method, and the segmentation algorithm is improved to make the algorithm run more efficiently. Considering the real-time problem, the lane extraction method is proposed to improve the real-time performance of the system. For the statistics of the fleet length and the number of vehicles, the row coordinate model is established, and the distance conversion between the image coordinate system and the actual coordinate system is realized. The length of the vehicle fleet is estimated, which provides an important parameter for the follow-up hardware system realization.
【學位授予單位】：內(nèi)蒙古大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TP391.41;U495

【參考文獻】

相關期刊論文 前1條

1 靳濤;張紅星;;基于動態(tài)圖像識別的智能交通燈控制[J];科技傳播;2012年20期

，

本文編號：2436731