Crew Activity Analyzer

Future space missions will require planning for human interactions in a confined and potentially harsh environment over extended periods of time. The study of human behavior in simulated environments is one means of planning for these missions. The proposed innovative Crew Activity Analyzer (CAA) will provide NASA with an efficient and cost-effective tool for automatically identifying patterns of group activity from image data of these simulated environments. The CAA will be more objective and accurate than using human observers to process the information from visual images. The CAA concept is based on computer vision algorithms developed by Foster-Miller. The tool will provide an automated means for determining patterns and timing of group activity. Human involvement in processing of the image data will be minimal. Phase I objectives are to 1) select and prioritize the group activities to be analyzed, 2) validate method for segmentation of image data, 3) develop suitable techniques for tracking individuals and 4) perform validation experiments, including computing metrics of individual and collaborative time and space usage. (P-020614)