A Robust Hough Transform-Based Track Initiation Method for Multiple Target Tracking in Dense Clutter

Abstract

Existing multiple track initiation methods based on Hough Transform have flaws, such as huge computation load in dense clutter environment and performance degradation in measurement error. To improve those defects, we propose a robust Hough Transform-based track initiation method for multiple target tracking in dense clutter. The whole proposed method is divided into three steps. The first step is to screen data through the grid-based velocity test. The second step is to cluster the screened data into groups based on eight connectivity principle of grids. The last step is to respectively make Hough Transform on each group data to achieve track initiation. The experiment results illustrate the proposed algorithm outperforms traditional algorithms on Hough transform based track initiation.