A Novel Visual Tracking Method Based on Moth-Flame Optimization Algorithm

摘要

Moth-flame optimization algorithm (MFO) is a new meta-heuristic optimization algorithm that mimics the motion of moths flight around the flames. In this work, a novel MFO-based visual tracking method is proposed by interaction between flames and moths. Firstly, visual tracking is expressed as searching for target in whole search space. Then, the spiral flight of moths is employed to enhance search capabilities. And the mechanism of reducing the number of flames gradually is applied for improving the convergence rate of the proposed algorithm. Finally, an appearance model based HOG feature is established to measure the similarity between the target and the candidate samples. In light of the best confidence value, the tracking target is located. Extensive experiments have demonstrated the effectiveness of proposed tracker.