Inspired by Human Eye: Vestibular Ocular Reflex Based Gimbal Camera Movement to Minimize Viewpoint Changes

Abstract

Human eyeballs move relative to the head, resulting in optimal changes in the viewpoint. We tested similar vestibular ocular reflex (VOR)-based movement on Zenmuse-X3 gimbal camera relative to pre-defined YAW movements of the DJI Matrice-100 unmanned aerial vehicle (UAV). Changes in viewpoint have various consequences for visual and graphical rendering. Therefore, this study investigated how to minimize these changes. OpenGL visualization was performed to simulate and measure viewpoint changes using the proposed VOR-based eyeball movement algorithm and compared with results of VOR based gimbal movement. The gimbal camera was setup to render images (scenes) on flat monitors. Positions of pre-fixed targets in the images were used to measure the viewpoint changes. The proposed approach could successfully control and significantly reduce the viewpoint changes and stabilize the image to improve visual tracking of targets on flat monitors. The proposed method can also be used to render real-time camera feed to a head-mounted display (HMD) in an ergonomically pleasing way.