Bioinspired Visuovestibular Artificial Perception System for Independent Motion Segmentation

In vision based systems used in mobile robotics and virtual reality systems the perception of self-motion and the structure of the environment is essential. Inertial and earth field magnetic pose sensors can provide valuable data about camera ego-motion, as well as absolute references for structure feature orientations. In this article we present several techniques running on a biologically inspired artificial system which attempts to recreate the “hardware” of biological visuovestibular systems resorting to computer vision and inertial-magnetic devices. More specifically, we explore the fusion of optical flow and stereo techniques with data from the inertial and magnetic sensors, enabling the depth flow segmentation of a moving observer. A depth map registration and motion segmentation method is proposed, and experimental results of stereo depth flow segmentation obtained from a moving robotic/artificial observer are presented.