The research currently conducted operates under the tenet of providing an increased level of situational awareness (SA) for soldiers on and off the battlefield using a combination of images and spatial information. More specifically in image and spatial data fusion, there exists a need to create a methodology to achieve near autonomous spatial and temporal tracking of entities, providing position and orientation (pose) data for use in tactical displays. Although image reconstruction has not been employed in the application of force tracking, this research is motivated by the positive results reconstruction has achieved in real-time tracking and depth estimation. Thus this study concentrates on the development of a precision force tracking framework by employing recursive structure from motion based algorithms to perform image reconstruction.;By capturing video streams mimicking those obtained from unmanned aerial vehicles (UAVs), the algorithm was able to provide refined three-dimensional (3-D) pose estimates of an observed entity. To calculate the error, the reconstructed points were then compared to the 3-D ground truth data. Extensive analysis illustrated the algorithm exhibited good performance when the scene contained feature points with little noise and when the relative motion was not exceedingly fast although quality results were also achieved when reconstructing datasets with higher levels of motion. More importantly, the system displayed good performance in accuracy, processing and uploading time, and reducing bandwidth consumption.;Lastly the approximated world coordinates addressed two integral topics. The first area provided answers to the questions when (temporal data) and where (spatial data), giving rise to the location of blue (friendly), red (enemy), and unknown forces at a specified time. The second topic tackled the issue of fusing data obtained from UAVs by decreasing the amount of bandwidth consumed when uploading vital information. This solution helped alleviate challenging demands placed on SA and UAV image processing when utilizing networks in a bandwidth-restricted environment. When integrated with computer vision as well as synthetic models emulating that of a battlefield, the findings of the precision force tracking framework will assist in the overall provision of tactical information and enhanced situational awareness to warfighters during combat.
展开▼
