Multitarget Classification and Estimation Using Clustering Techniques

A modular approach to solving the multisensor-multitarget problem in space is presented. This approach combines the clustering techniques developed in the field of pattern recognition with the filtering techniques developed to estimate the states of several targets in space. Results from the clustering analysis are presented, and their relevance to the identification of targets and data fusion are described using two scenarios having multiple targets and using orbiting platforms as sensors. The estimation error histories of positions and velocities are also given.