A mapping method for 3D satellite and sensor images using a road extraction algorithm for occlusion processing of virtual targets

Abstract

Augmented reality is a challenging issue in computer-based training (CBT) when a previous training process that required a large amount of resources is changed to a training simulation technique. This paper describes the results of the research and development of algorithms that are based on a virtual target display on a real CCD image with a specific scenario for a realistic training simulation. We created a realistic 3D model with a high resolution geographic tag image file format (GeoTIFF) satellite image and digital terrain elevation data (DTED), and we extracted the road area from a given sensor image with an existing and enhanced snake algorithm for the occlusion processing. We also propose a moving synchronisation technique that projects the target onto the sensor image according to the marked moving path on a 3D satellite image by applying a thin-plate-spline (TPS) interpolation function, which is an image warping function, on the two given sets of the corresponding control point pair. The developed algorithms and the implementation results are described.