Abstract
Aiming at tracking performance of real-time and robustness of crane monitoring system, current target tracking methods are contrasted in this paper, such as Tracking-Learning-Detection (TLD) algorithm, MeanShift and particle filter. Experimental results show that MeanShift and particle filter algorithm have the advantage over TLD in real-time performance. However, TLD can adapt to the crane transformation with preferable self-learning capability. And TLD enables to re-track the object even if the object is sheltered or missed momently. Therefore, the crane monitoring system based on TLD algorithm can not only detect and track the target accurately but also be superior to another two algorithms in performance of robustness.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
