Deep learning-powered image sensor to detect overcrowding in COVID-19-infected cities
Crowded places tend to become a hub for transmission of infectious diseases. The COVID-19 pandemic has shown us that we need to find ways to manage crowded areas to help limit the spread of infectious diseases. Unmanned aerial vehicles (UAVs), such as drones, can detect and record environmental conditions at different heights above the ground in real time. This makes them ideal for detecting overcrowding and unusual mob behavior, such as riots.
To achieve this, a team of scientists led by Professor Gwanggil Jeon from Incheon National University, South Korea, has developed a real-time image sensing system using algorithms. deep learning. “In this paper, we propose a real-time system to detect overcrowding and abnormal behavior of crowds. The monitoring system detects overcrowding by using UAVs communicating with social network. monitoring system (SMS),” said Professor Jeon.
Their findings were made available in IEEE Transactions on Industrial Informatics.
The system can be broken down as follows. First, the UAV recorded the crowd. The video frames from this footage are then fed into the decision-making system. In this decision-making system, features are first extracted using a ‘modified ResNet architecture’. The features are then selected using the “water cycle algorithm” (WCA) and then classified into different categories that describe crowding or crowd behaviour. Finally, this data is put into an SMS message.
The proposed model was able to successfully detect overload with 96.55% accuracy in real time. It can also detect crowd behavior, which is important for tracking and suggesting alternative routes to prevent the spread of infectious diseases. Furthermore, the system is more robust and provides fast detection with high accuracy thanks to the modified ResNet architecture, which has fewer front-end layers.
“Our new system can be deployed and implemented in Smart Cities to help fulfill some purpose of the social system. It is a powerful tool that can help limit the spread of infectious diseases such as COVID-19 by monitoring crowd behavior and recommending appropriate crowd movement routes by SMS,” concluded Professor Jeon.
The proposed system paves the way for real-world application of vision sensors for crowd control purposes.
Khosro Rezaee et al., Intelligent image sensors for overcrowding in COVID-19 infected cities using modified deep transformation learning, IEEE Transactions on Industrial Informatics (2022). DOI: 10.1109/TII.2022.3174160
Provided by Incheon National University
quote: Deep learning-powered image sensor for overcrowding detection in COVID-19-infected cities (2023, February 7) retrieved February 7, 2023 from https://techxplore .com/news/2023-02-deep-learning-assisted-visual-overcrowding -covid-.html
This document is the subject for the collection of authors. Other than any fair dealing for private learning or research purposes, no part may be reproduced without written permission. The content provided is for informational purposes only.