Comparison of Different Level Fusion Schemes for Infrared-Visible Object Tracking: An Experimental Survey

Infrared and visible image fusion tracking is now revealing its great potential and breeding a qualitative leap in visual object tracking field. Valuable complementary information provided by infrared and visible images is the key of improving tracking performance and plays a critical role when facing multiple complex challenges. In this paper, a summary for datasets of infrared and visible image sequences is intentionally made to conclude our preparatory work for building a standard benchmark. Besides, a general framework of fusion tracking on three different levels is proposed to conclude the fusion tracking schemes for the first time, which can generally include all the fusion tracking algorithms. Extensive experiments under complex scenarios have been conducted to evaluate the performances of the three fusion tracking schemes. The results show the significance of the fusion tracking on infrared and visible image sequences. Finally, this survey proactively concludes future research directions and potential improvements in this field.