Optical Flow based person following behaviour of a robot

摘要: Tracking features between two consecutive images captures the essence of motion in order to categorize objects (either static or moving) scene. There has been a lot literature on tracking (sparse dense) and improvements have also proposed over time. Many these methods try extract either through global optic flow methods, Horn-Schunck local LucasKanade. The analysis is more scenes taken from camera which background remains stationary but it becomes challenging moving as inherited into objects. We examine problem tailing single well multiple people mounted mobile robot present solution for same. In method, an alternative approach computation by formulating energy minimization framework. computed field filtered using spatial relative velocity based filter determine potential Color depth information then used finally segment correctly classify works different testing environments including change illumination, presence many textured similar color.