Clustering Gaussian mixture reduction algorithm based on fuzzy adaptive resonance theory for extended target tracking
Yongquan Zhang, Hongbing Ji · IET Radar Sonar & Navigation · 2014
This study presents a global Gaussian mixture reduction (GMR) algorithm via clustering, which is based on a fuzzy adaptive resonance theory (FART) neural network architecture. Therefore the authors call the proposed algorithm as GMR based on the fuzzy ART (GMR‐FART) in this study. The architecture of GMR‐FART is similar to that of the FART, however, its choice function, match function and learning update equations are characterised by features of Gaussian mixture (GM). The proposed algorithm automatically forms categories (i.e. the reduced GM components) via a feedback mechanism. The performance of GMR‐FART is evaluated by the normalised integrated squared distance measure which describes the deviation between the original and the reduced GM. The proposed algorithm is tested on both one‐dimensional (1D) and 4D simulation examples, and the results show that the proposed algorithm can accurately approximate the original mixture and requires less computational burden, and is useful in extended target tracking.