Shared more. Cited more. Safe forever.
    • advanced search
    • submit works
    • about
    • help
    • contact us
    • login
    View Item 
    •   MOspace Home
    • University of Missouri-Columbia
    • Graduate School - MU Theses and Dissertations (MU)
    • Theses and Dissertations (MU)
    • Theses (MU)
    • 2019 Theses (MU)
    • 2019 MU theses - Freely available online
    • View Item
    •   MOspace Home
    • University of Missouri-Columbia
    • Graduate School - MU Theses and Dissertations (MU)
    • Theses and Dissertations (MU)
    • Theses (MU)
    • 2019 Theses (MU)
    • 2019 MU theses - Freely available online
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.
    advanced searchsubmit worksabouthelpcontact us

    Browse

    All of MOspaceCommunities & CollectionsDate IssuedAuthor/ContributorTitleIdentifierThesis DepartmentThesis AdvisorThesis SemesterThis CollectionDate IssuedAuthor/ContributorTitleIdentifierThesis DepartmentThesis AdvisorThesis Semester

    Statistics

    Most Popular ItemsStatistics by CountryMost Popular AuthorsStatistics by Referrer

    Artificial intelligence enabled automatic traffic monitoring system

    Mandal, Vishal
    View/Open
    [PDF] MandalVishalResearch.pdf (2.052Mb)
    Date
    2019
    Format
    Thesis
    Metadata
    [+] Show full item record
    Abstract
    The rapid advancement in the field of machine learning and high-performance computing have highly augmented the scope of video-based traffic monitoring systems. In this study, an automatic traffic monitoring system is proposed that deploys several state-of-the-art deep learning algorithms based on the nature of traffic operation. Taking advantage of a large database of annotated video surveillance data, deep learning-based models are trained to track congestion, detect traffic anomalies and tabulate vehicle counts. To monitor traffic queues, this study implements a Mask region-based convolutional neural network (Mask R-CNN) that predicts congestion using pixel-level segmentation masks on classified regions of interest. Similarly, the model was used to accurately extract traffic queue-related information from infrastructure mounted video cameras. The use of infrastructure-mounted CCTV cameras for traffic anomaly detection and verification is further explored. Initially, a convolutional neural network model based on you only look once (YOLO), a popular deep learning framework for object detection and classification is deployed. The following identification model, together with a multi-object tracking system (based on intersection over union -- IOU) is used to search for and scrutinize various traffic scenes for possible anomalies. Likewise, several experiments were conducted to fine-tune the system's robustness in different environmental and traffic conditions. Some of the techniques such as bounding box suppression and adaptive thresholding were used to reduce false alarm rates and refine the robustness of the methodology developed. At each stage of our developments, a comparative analysis is conducted to evaluate the strengths and limitations of the proposed approach. Likewise, IOU tracker coupled with YOLO was used to automatically count the number of vehicles whose accuracy was later compared with a manual counting technique from CCTV video feeds. Overall, the proposed system is evaluated based on F1 and S3 performance metrics. The outcome of this study could be seamlessly integrated into traffic system such as smart traffic surveillance system, traffic volume estimation system, smart work zone management systems, etc.
    URI
    https://hdl.handle.net/10355/72268
    Degree
    M.S.
    Thesis Department
    Civil and Environmental Engineering (MU)
    Rights
    OpenAccess.
    This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License.
    Collections
    • 2019 MU theses - Freely available online
    • Civil and Environmental Engineering electronic theses and dissertations (MU)

    Send Feedback
    hosted by University of Missouri Library Systems
     

     


    Send Feedback
    hosted by University of Missouri Library Systems