Some aspects of optimal human-computer symbiosis in multisensor geospatial data fusion

Nowadays vast amount of the available geospatial data provides additional opportunities for the targeting accuracy increase due to possibility of geospatial data fusion. One of the most obvious operations is determining of the targets 3D shapes and geospatial positions based on overlapped 2D imagery and sensor modeling. 3D models allows for the extraction of such information about targets, which cannot be measured directly based on single non-fused imagery. Paper describes ongoing research effort at Michigan Tech attempting to combine advantages of human analysts and computer automated processing for efficient human computer symbiosis for geospatial data fusion. Specifically, capabilities provided by integration into geospatial targeting interfaces novel human-computer interaction method such as eye-tracking and EEG was explored. Paper describes research performed and results in more details.