Personnel Detection at a Border Crossing—An Exercise

Abstract : In March 2012, an international team of scientists, engineers, and technicians gathered at the southwest border of the United States with specialized equipment to collect data on people, animals, and vehicles travelling in the rugged terrain. The goal of the effort is to collect data in the natural environment and develop robust algorithms to detect people, animals, and vehicles with fewer false alarms and high confidence. The Canadian team used SASNet; the Israeli team used Pearls of Wisdom; University of Memphis brought a Profiling sensor; and the University of Mississippi, Night Vision and Electronic Sensors Directorate, the Space & Naval Warfare Systems Command (SPAWAR), and U.S. Army Research Laboratory brought their equipment to collect the data. Representatives from Finnish Defense participated in observing the team. Some of the sensor modalities used are acoustic, seismic, passive infrared (IR), profiling sensor, sonar, and visible and IR imaging sensors. Some description of the sensors and their data analysis is presented. In this report, we present the data collection effort and some of the algorithms developed for various sensor modalities along with the results on the field data.