An Algorithmic Approach for Real Time People Counting with Moving Background

Abstract

The research paper implements the concept of real-time people counting. Real time people counting is a relevant source of information in various applications like security or people management, such as in shopping malls, footfall or fleet management in public transport, congestion management and tourist flow estimation. Today, a major use of people counting is done there where we have a cluttered audience and hence we have to make the management of any kind. The proposed algorithm does real time people counting using mean shift with deep learning data sets. The camera used in the system takes the video sequence as an input to the algorithm. The various steps of the algorithm then detect whether the object present in the video is a human or not. After the object is identified as a human, the counting algorithm is then applied to find the number of people in the video and display the counted result.