Moving Objects Tracking Method Based on Discharged Optical Flow
Ivan Aleksandrovich Vozvakhov, Mikhail Dmitrievich Strakhov, Vladislav Aleksandrovich Ershov, Marat Yuryevich Agapov, Viktor Viktorovich Tatarinov, Anastasiia Pavlovna Rozhok · 2022 4th International Youth Conference on Radio Electronics, Electrical and Power Engineering (REEPE) · 2022
This study presents the tracking moving object method based on the Lucas-Kanade algorithm with feature points regeneration. The given algorithm provides the time-based ability to scale feature points of a moving object, which reduces the probability of tracking errors. Besides, this method allows to track objects, which have parts with different motion velocities. The Lucas-Kanade method and the tracking method based on the biological regeneration principle were considered in our research. Based on the reviewed methods, a new moving object tracking method with adaptive change in the number of feature points was proposed. The method allows reducing the probability of losing the tracked object when passing through the occlusion area.