Scholarlink

The traffic management system in Davao City faces challenges with frequent downtimes of traffic lights, leading to subpar performance. This study aimed to address these issues by developing an Intelligent Traffic Signaling System (ITSS), offering an alternative approach to the city’s current traffic management strategies. This ITSS employs the random forest algorithm (RFA) which aimed to determine the most efficient traffic light switching periods based on parameters such as vehicle count, time of day, and the identification of emergency vehicles. Additionally, YOLOv5 is utilized for accurate traffic counts and emergency vehicle identification. Training the model consists of using the concepts of Queuing Theory and Traffic Intensity using existing traffic surveillance footage from preexisting infrastructure. The RFA has a testing accuracy of 95.39% where the confidence level threshold of YOLOv5 was 75%. This research represents a shift towards a more technologically advanced traffic management system, potentially offering improvements in traffic flow and efficiency in Davao City by simply using preexisting infrastructure which saves time and resources.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Intelligent Traffic Signaling System for a Congested Four-Way Intersection Through Random Forest Algorithm and YOLOv5

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Uean Carlo T. Bacsal,
  • James Raye C. Camaganacan,
  • Jerrik Michael V. Estardo,
  • John Paul T. Cruz,
  • Jay T. Cabuñas,
  • Michael G. Calamba

摘要

The traffic management system in Davao City faces challenges with frequent downtimes of traffic lights, leading to subpar performance. This study aimed to address these issues by developing an Intelligent Traffic Signaling System (ITSS), offering an alternative approach to the city’s current traffic management strategies. This ITSS employs the random forest algorithm (RFA) which aimed to determine the most efficient traffic light switching periods based on parameters such as vehicle count, time of day, and the identification of emergency vehicles. Additionally, YOLOv5 is utilized for accurate traffic counts and emergency vehicle identification. Training the model consists of using the concepts of Queuing Theory and Traffic Intensity using existing traffic surveillance footage from preexisting infrastructure. The RFA has a testing accuracy of 95.39% where the confidence level threshold of YOLOv5 was 75%. This research represents a shift towards a more technologically advanced traffic management system, potentially offering improvements in traffic flow and efficiency in Davao City by simply using preexisting infrastructure which saves time and resources.