OMNI DIRECTIONAL MOVING OBJECT DETECTION AND TRACKING WITH VIRTUAL REALITY FEEDBACK

摘要

Computer vision methods are commonly used to detect and track motion using conventional cameras, however, that is limitedwith the field of view (FOV) of the camera. This study is to attempt to overcome this challenge by using a 360 degree camera. Our approach utilizes background subtracter from OpenCV Library which creates a continuously updating background model for the motion detection. The model is subtracted from the current frame leaving contours symbolizing the movement observed in the camera view. These contours are then analyzed and processed so that the system can track the largest contour. The tracked movement is outlined and directed to the user via Virtual Reality (VR) headset. The VR headset only displays a 60 degree portion of the camera view to the user which provides more realistic situational awareness of the surroundings for the user. These activities are a part of a larger effort to establish a foundation for autonomous unmanned vehicle systems with situational awareness capabilities.