A robust moving object tracking

This paper extends the work in [4] by introducing the neighborhood querying process for reducing the effects of sensor observation errors. The objective of this work is to track moving objects by using one-bit information from sensor nodes that provide information whether an object is moving towards a group of sensors or moving away. In [3], they allowed that sensor nodes make observation errors but did not introduce the methodology to handle this problem. In this work, we use the neighborhood query process to mitigate the effects of observation errors. The process is based on neighboring information that can be used to confirm whether the current observed data are correct or not. After that, it is the responsibility of the support vector machine to classify groups of data and particle filters to track objects that follow the direction of the vector perpendicular to the optimal hyper plane obtained by the support vector machine.