The leaky integrate-and-fire neuron model for a rigid and a non-rigid object tracking

Spiking neural networks (SNNs) fall into the third generation of artificial neural network models, increasing the level of realism in a neural simulation. In this paper, a spiking neural network is presented for detecting and tracking of a moving object in video sequences with a static camera. The motion estimation of the object is carried out by minimizing a Hausdorff distance measure. The system has been successfully tested with various real video sequences. The results showed that our system can track the identified target over subsequent video frames even in occlusion case.