ABSTRACT
Crowd density management in the transport sector is still one of the ongoing research problems. Intelligent Transportation System (ITS) is one of the branches of smart cities that aim to achieve better traffic efficiency. The intelligent transport system optimizes the traffic congestion control by acquiring real time data. Optimized traffic congestion control demands a robust system that could count the number of people inside a carrier for taking optimized decisions. In this paper we proposed an intelligent algorithm named Modified Intelligent Centroid Tracker and Counter (MICTC) that could detect, count, and measure the distance between humans in a closed and controlled environment. The proposed algorithm is vision based and the scope of the work is to optimize the congestion control inside the passenger carrier and supports countless use cases like smart transport, buildings, and other demography where social distancing is enforced. MICTC algorithm not only offers visual indication with a bounding box but also generates metadata which gives a clear picture to the concerned operational or administrative head regarding the current passenger count status. The work deployed in the public transport sector as a candid spot to operate. The algorithm delivers an adequate transport facility to the public, as it gathers information on crowd density in a public transport medium to the commuters of every region. The work gathers crowd density information and provides commuters a suggestion on availability of seats in the carrier, which then saves time, avoids catching the crowded carriers, ensures social distancing, and standardizes the public transportation system which has practical significance. On experimental analysis we could infer that the proposed approach works with accuracy of 0.81,0.83, 0.85, 0.88, 0.82, 0.89 on VISOR, Kaggle, CALTECH, Penn-Fudan, Daimler Mono and INRIA respectively.