Extended Target Measurement Set Partition Algorithm Based on KFCM Clustering

Aiming to the problem of measurement set partition of multiple extended target, a measurement set partition algorithm based on kernel fuzzy C-means (KFCM) clustering is proposed. Firstly, according to the FCM algorithm, the initial cluster centers are generated. Then, the distance function of FCM algorithm are changed by kernel function. Finally, this paper constructs new interation cycle to obtain tracking results. Experimental results shows that the proposed algorithm improves the performance of multiple extended target tracking compared with FCM partition and K-means++ partition effectively.