Implementing Real-time Visitor Counter Using Surveillance Video and MobileNet-SSD Object Detection: The Best Practice

Authors

Nasser Al Musalhi, Department of Information Technology, University of Technology and Applied Science, Ibra, Oman.Follow
Ali Mohammed Al Wahaibi, Faculty of Communication, Visual Art and Computing, Universiti Selangor (UNISEL), Bestari Jaya, Malaysia.Follow
Mohammed Abbas, Department of Information Technology, University of Technology and Applied Science, Sur, Oman.Follow

Abstract

Counters that keep track of the number of people who enter a building are a useful management tool for keeping everyone who uses it safe and happy. This paper aims to employ the MobileNet-SSD machine learning approach to implement a best practice for visitor counter. The researchers have to build a different scenario test dataset along with the MOT20 dataset to achieve the proposed methodology. Implementing different experiments in single-user, one-one; two-two users; many-two, and multiple users in different walking directions to detect and count shows varied results based on the experiment type. The best achieved by single-user and one-to-one model; both are scored 100% of detecting and calculating for in or out.

Keywords

CNN, Mobilenet-SSD, MobileNet, Object tracking, real-time Object detection, SSD

Article Type

Special Issue Article

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

How to Cite this Article

Al Musalhi, Nasser; Al Wahaibi, Ali Mohammed; and Abbas, Mohammed (2024) "Implementing Real-time Visitor Counter Using Surveillance Video and MobileNet-SSD Object Detection: The Best Practice," Baghdad Science Journal: Vol. 21: Iss. 5, Article 28.
DOI: https://doi.org/10.21123/bsj.2024.10540