Research News

Manish Kumar | Department of Mechanical and Materials Engineering

Simultaneously tracking multiple targets is a problem of interest for both civilian and military applications. The problem becomes increasingly difficult when the number of sensors greater than or equal to 3, as the S-D assignment algorithm for track-to-track correlation has been shown to be NP-Hard. In small scenarios, solutions can be achieved in relatively fast times. However, as the number of sensors and targets increase, standard approaches fail to scale for real-time execution. This paper shows that near-optimal results can be achieved in real-time by decomposing the problem space by fuzzy c-means clustering, then exploiting the partition matrix to assist in cluster element modifications. Furthermore, the paper explores the correct number of clusters to utilize when the true number of targets in the environment is unknown.


Select for more information about research work of Dr. Manish Kumar.


Cited from Dr. Kumar's Publication, "Generic Clustering Approach to Track-to-Track Correlation for Multisensor-Multitarget Environments"


Select for more information about research activities within the College of Engineering and Applied Science.