An Integrated Object Tracking and Covert Visual MIMO Communication Service for Museum Security System Using Single Vision Sensor

Abstract

This paper explores the spatial coverage potential of a vision sensor to implement a dual service for a museum security system. Services include object monitoring and invisible visual Multiple-Input and Multiple-Output (MIMO) communication. For object monitoring, the Kalman filter algorithm is used. To implement a secret visual MIMO communication, an Adaptive Row-Column (ARC)-based LSB substitution technique is used along with the Integer Wavelet Transform method. These proposed services are implemented using existing electronic visual displays and surveillance cameras. Experiments are carried out using a sample object and four stego images. The performance of the object tracking system is measured in terms of accuracy by comparing the actual and estimated position of an object. Similarly, the performance of the covert visual MIMO system is measured at two different stages. On the transmitter side, differences between the cover and stego-images are used to measure imperceptibility. Likewise, at the receiver, differences between the original and the reconstructed data images are used to measure the accuracy. Results illustrate that the proposed system has improved imperceptibility, accuracy, and robustness, as compared to existing techniques.