The system is designed to operate on portable devices with limited energy resources and processing power. Users interact with a three-dimensional (3D) model of the region of interest. Each user can determine their position in terms of latitude, longitude, and elevation through a direct connection to GPS satellites.

The system enables users to navigate the terrain and place landmarks based on their observations in the field. The virtual 3D terrain can be rendered transparently, allowing users to view landmarks that would otherwise be obscured by terrain features or obstacles. Whenever a network connection is available, all landmark data can be synchronized and shared with other users through the central coordination system.

Users can mark assets by selecting different icons and placing them on the map using the device's touch screen. Smooth and responsive movement throughout the virtual environment is achieved using optimized graphics rendering algorithms.

The system also provides a distance measurement tool that allows users to calculate the straight-line (bird's-eye) distance between two points by simply dragging a pointer across the touch screen.

During the course of this study, we developed two versions of a reference application that served as a testbed for evaluating the proposed system.

The first version was completed in February 2007 and provided GPS-based positioning along with 3D virtual terrain navigation capabilities. It utilized a low-resolution 3D terrain model, enabling it to run efficiently on mobile phones equipped with 3D graphics support.

The second version represents the current implementation and incorporates our optimized 3D rendering and landmarking algorithms, offering improved performance and enhanced functionality.

The application was developed using Microsoft Visual Studio 2005, the C# programming language, the .NET Compact Framework for mobile devices, and DirectX Mobile for 3D graphics processing. Testing was conducted on touchscreen-enabled mobile devices to facilitate intuitive user interaction and improve the flexibility of navigation and landmark placement.

Considering a surveillance team as the primary users of the system, the application enables users to place and manage multiple landmarks within a specified operational area. To support this functionality, the system implements a 2D-to-3D unprojection algorithm that determines the three-dimensional coordinates corresponding to a point selected on the two-dimensional touchscreen. Whenever a user touches the display, the system computes the corresponding 3D coordinate, provided that the selected point lies on the terrain surface.

The current implementation employs a proprietary terrain model rather than a standard Digital Elevation Model (DEM). Ongoing research focuses on supporting Digital Terrain Elevation Data (DTED) formats for deployment on handheld devices.

Because DTED Level 1 and higher datasets exceed the memory capacity of the target mobile devices, we are investigating approximation algorithms that estimate elevations for intermediate points using DTED Level 0 data. This approach aims to provide higher-quality terrain representation while remaining within the memory and processing constraints of resource-limited handheld platforms.

To further improve system performance, future development will focus on migrating the implementation to a lower-level programming language to reduce the overhead introduced by the underlying software framework and to enable more efficient 3D graphics processing.

The project has been developed primarily by a team of undergraduate and graduate students, each contributing on a part-time basis throughout the course of the research.