A Hybrid Multimodal Tracking System for boarder surveillance

Video surveillance activity is the act of observing the behavior of the person/object under surveillance. Surveillance at borders is a very sensitive issue of security. However, no real time automated border surveillance system has been developed yet. The proposed algorithm addresses the large distance between object and camera problem, due to which visual data can capture very limited scene information at borders, by analyzing additional seismic signals generated by the object while moving on the ground. If there are two objects, a unique pattern is extracted from the received mixed seismic signal for object identification. And Particle Swarm Optimization (PSO) algorithm is used to track the object. For experimental evaluation video samples with seismic signals are recorded which proves that integration of seismic signal analysis helps to identify the suspicious object on border which is not visible in images and has reduced the RMSE in predicting the new position of the object.