REAL TIME OBJECT MONITORING USING ADAPTIVE CORRELATION FILTER

Abstract

Real time localization of a moving or stationary object in a video stream is one of the most complex tasks in video processing due to changing background conditions or full / partial object overlaps is considered. For present moment, a number of techniques and solution proposed, nevertheless the task is still actual and no one can speak about revolutionary breakthrough. Present article describes real-time object tracking algorithm based on adaptive correlation filter usage. The algorithm is autonomous and doesn't need parameters adjustments from the beginning till the end of its functioning.

Keywords

• автосопровождение объектов
• адаптивный корреляционный фильтр