Relaxed Phase Optimization for 0-1 Combinational Problem in MHT

Multiple Hypothesis Tracking (MHT) is a key algorithm in the field of target tracking, especially for dense multi-target tracking in strong clutter background. This paper proposes a new method based on phase optimization to solve the global hypothesis generation problem in MHT. By converting the combinatorial optimization problem into a phase optimization problem and solving it using the alternating direction method of multipliers (ADMM), we effectively solve the contradiction between computational complexity and optimization accuracy. Experimental results show that this method shows excellent performance in multiple complex scenarios.