Augmented reality – 3D objects in context – Part 1
aero3Dpro technology has created a large warehouse of 3D models for us internally. So we have been working on a few presentation aspects, bringing the models to web is possible via WebGL in the right browsers and we can of course print them using a decent 3D printer. The other very interesting way to present the models and the value they add to traditional orthophotography is using Augmented Reality. Using the camera in the now common place smartphones and tablets you can overlay the rich 3D models we have been collecting over any aerial image of the similar area.
Augmented reality embeds the rich city models captured using photogrammetry into the overall context, allowing the planner, architects and developers to get an understanding of the environs of their project and change details. Augmented reality on a coarse scale can be performed using GPS and the modern GPS free location services using signals of opportunity. The Google Ingress game is based on this idea. To augment the view right in front of you detailed knowledge of the 3D objects is needed, providing the depth cues using perspective, occlusion, shadowing and depth-of-field blurring to merge new information into the 2D video.
(Click to enlarge)
One of the leaders in augmented reality applications is Total Immersion, the typical specialised applications are markerless tracking and face tracking. With faster CPU’s becoming available on mobile platforms, feature tracking and on the go augmented reality. To squeeze the best performance out of the processors vision toolkits are typically sponsored by the chip manufacturers, Intel started the trend with OpenCV, with the ARM architecture gaining dominance on the mobile platform, chip vendors such as Qualcomm have taken up the ARM optimised vision and tracking toolkit challenge.
For photogrammetry and GIS applications augmented reality allows the presentation of 2.5 D (Textured DSM) or True 3D data (aero3Dpro models) to be presented in context atop orthophotography or topographic maps. This allows the best of both worlds approach to viewing geographic data, adding a new dimension to existing flat maps. Navigation around the model is very natural and intuitive, just short of actually being there.