Density Clustering Based Data Association Approach for Tracking Multiple Target in Cluttered Environment

Abstract

Tracking of multiple targets in heavy cluttered environments is a big challenge. One usual approach to overcome this problem is using data association process. In this study, a novel fuzzy data association based on density clustering for multi-target tracking is proposed. In the proposed algorithm, the density clustering approach is used to cluster the measured data points. This approach is used instead of gates to eliminate false alarms that originate from invalid measurements. Then the association weights of the validated measurements are determined based on the maximum entropy fuzzy clustering principle. The efficiency and effectiveness of the proposed algorithm are compared with JPDAF, MEF-JPDAF and Fuzzy-GA. The results demonstrate the main advantages of the proposed algorithm, such as its simplicity and suitability for real-time applications in cluttered environments.