Two-stage filtering based method for multiple target tracking

A method for detecting and tracking an unknown number and time-varying number of targets within a surveillance area, the method comprising:

• (a) providing a first non-linear filter for detecting at least one target within the area, wherein the first filter is initialized by a uniform density covering the area, and is reset to said uniform density upon a pre-determined number of iterations,
• (b) providing a clustering technique to detect clusters corresponding to the at least one target, wherein the clustering technique is adapted to operate at said pre-determined iterations, and before said first filter is reset,
• (c) providing a second non-linear filter for tracking the at least one target detected by the first filter which is represented by at least one cluster detected by the clustering technique,
• wherein the first non-linear filter and the clustering technique also operate to detect the absence of targets within the surveillance area.