English  |  正體中文  |  简体中文  |  Items with full text/Total items : 90453/105672 (86%)
Visitors : 11996192      Online Users : 346
RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    ASIA unversity > 資訊學院 > 資訊傳播學系 > 博碩士論文 >  Item 310904400/112476


    Please use this identifier to cite or link to this item: http://asiair.asia.edu.tw/ir/handle/310904400/112476


    Title: 透過卷積神經網路判斷公路行車流量
    Identification of Traffic Flow Using Convolutional Neural Networks
    Authors: 黃, 予
    HUANG, YU
    Contributors: 資訊傳播學系
    Keywords: 影像辨識;人工智慧;深度學習;卷積神經網路;行車流量偵測
    Image recognition;Artificial intelligence;Deep learning;Convolutional neural network;Traffic flow detection
    Date: 2019
    Issue Date: 2019-11-11 11:46:50 (UTC+8)
    Publisher: 亞洲大學
    Abstract: 臺灣地區運輸路網發達、車輛普及化,進而衍生許多交通相關問題,而交通壅塞更成政府施政的重點之一。如何開發有效而可靠的自動分析方式於高速公路行車流量的判斷,作為各個交流道閘道儀控管制的參考依據,是一項極為重要而迫切的研究工作。本文將使用兩種不同的卷積神經網路(Convolutional Neural Network, CNN)來進行公路行車流量的判斷,達到自動判斷高速公路行車流量之目的。第一個卷積神經網路用於學習、處理與分割每張影像中的車道,並將所指定車道劃分出來,避免非車道影像影響辨識率。第二個卷積神經網路則用於判斷與辨認經由第一個神經網路處理後影像的行車流量。本文辨認行車流量的結果依照Google針對型車流量所使用的分類將其分為三類,分別為「壅塞」、「車多」、「通暢」。經由實驗結果證明,使用本文提出方法於判斷公路行車流量影像的辨識成功率最高可達92.5%,所以本文所提出的方法確實可以準確的判斷高速公路的行車流量。
    Traffic jam is a major problem in Taiwan. Developing a reliable analysis method for the judgment of expressway traffic flow is an important task. The analysis results can be utilized to effectively control the number of cars entering into expressway for each gateway. In this thesis, two Convolutional Neural Networks (CNN) will be used to define the traffic flow around one segment on the expressway. The first CNN is utilized to segment the lanes in an image. The second CNN is used to determine and identify the traffic flow of the image. Identified traffic flow is classified into three categories according to the classification rule used by Google, namely "jammed", "car more" and "smooth". The experimental results show that the proposed method can be used to define the traffic flow effectively by only analyzing images. The accuracy rate can reach 92.5%.
    Appears in Collections:[資訊傳播學系] 博碩士論文

    Files in This Item:

    File Description SizeFormat
    index.html0KbHTML67View/Open


    All items in ASIAIR are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback