Secure Occupancy Monitoring System for IoT Using Lightweight Intertwining Logistic Map

Vision-based occupancy counters have a wide range of applications in resource management, queue management, and handling emergency situations. In real time scenarios, occupancy counters are connected to the Internet. However, the Internet is an insecure channel and always prone to various types of attacks. When wired or Wireless Multimedia Surveillance Networks (WMSNs) are connected to the Internet, the privacy of innocent people can also be compromised. This paper presents a secure real-time scheme for IoT systems by intelligent integration of occupancy monitoring and chaos-based lightweight image encryption. Firstly, the real-time video was used to extract video frames through the single overhead camera. When people are detected in a frame, the intelligent system encrypts the current frame and also counts the people in/out and send occupancy count information to the cloud computing platform (ThingSpeak). Several experiments and security analyses were conducted to prove the security of partially encrypted video frames. Entropy, correlation coefficient, keyspace, number of pixel change rate, unified average change intensity, and histogram analysis prove the security of the proposed video encryption scheme. Furthermore, the proposed chaos-based encryption has faster processing time and hence the scheme can provide sufficient security in low power IoT devices.