An augmented reality (AR) object positioning system receives a target coordinate for a location of interest within a physical environment. The AR object positioning system then determines an AR device coordinate of an AR device within the physical environment and orientation information of the AR device at the AR device coordinate. The AR object positioning system calculates a plurality of image frame coordinates of an image frame based on the AR device coordinate and the orientation information when the image frame was captured. The AR object positioning system, in response to an image frame coordinate of the plurality of image frame coordinates corresponding with the target coordinate, renders on a display of the AR device, first annotation content relative to the image frame coordinate according to rendering instructions for the first annotation content.