Efficient Implementation Of Intelligent Traffic Management System Under Severe Adverse Weather Condition (aITMS)

Abstract

This research endeavor provides the groundwork for addressing the adversarial weather traffic events induced by meteorological phenomena such as rain haze or smog. The identifying and classifying automobiles are the essential parts of managing the traffic system. Setting up a mechanism to identify the automobiles in the pictures and alert the appropriate signaling system is crucial. According to the psychology of human learning humans are the finest planners because they manage and modify their tactics for their environment and develop cognitive learning skills by means their experiences. In the research we create a simulation from scratch and we observe natural detrimental effects for traffic signals and automobiles. Automobiles are labelled as cars bicycles buses and trucks to provide a more precise estimate. We employ objection detection techniques such as YOLOv4 to identify the number of automobiles in each location. Object recognition stands out as one of the most prevalent and widely employed applications of deep computational learning. The algorithm discussed is an intelligent convolution neuronal system enabling precise object recognition in images and videos with the capability to accurately delineate and restrict identified items. In the future our proposed approach may integrate improved human learning principles to a variety of practical scenarios granting swifter and more efficient insights.

Keywords

• Traffic
• CNN
• Weights
• Tensorflow
• Adverse Environment