A method and apparatus for automated assignment of geodetic coordinates to pixels of images of aerial video is disclosed. The method may include estimating a camera model for a first video frame of the aerial video, registering the first video frame to a Lidar DEM (digital elevation model) using the estimated camera model of the first video frame, iteratively refining the estimated camera model for the first video frame by composing the estimated camera model for the first video frame with the Lidar DEM registration, estimating a camera model for a next video frame of the aerial video by registering the next video frame to a previous video frame and composing the camera model of the next video frame with the previous video frame, and iterating for subsequent video frames.