On Sunday, 24 August 2008, in the midst of a celebration outside a house in the Abu Ghraib area of Baghdad, a young woman, who had just cleared a checkpoint, joins the celebration and blows herself up, killing 25 people and injuring 28. Individuals who carry bombs on their bodies and detonate those bombs in public places are clearly a force protection problem, all too familiar in the Middle East and with the real potential for such scenarios increasing in the West.

While vehicle bombs and IEDs are the primary method currently used to spread terror in Iraq and attack U.S. convoys, U.S. warfighters are starting to face suicide bombers. What is necessary is an inexpensive suicide bomber vetting system that can be used to screen individuals approaching a checkpoint while the individual is still 500 to 1,000 feet from the checkpoint. Gait variance may provide tipoff. Center for Technical Intelligence Studies and Research (CTISR) at the Air Force Institute of Technology, in collaboration with Dr. Charmaine Gilbreath of Defense Intelligence Agency’s Signature Support Office, Dr. Kathleen M. Robinette’s 711 Human Performance Wing Biomechanics Branch researchers (Julia B. Parakkat and Adam M. Fullenkamp), and Dr. Kimberly D. Kendricks, Assistant Professor of Mathematics at Central State University in Wilberforce, Ohio, are investigating video derived gait characteristics that are associated with an individual carrying a bomb on his or her body. The biometric question is whether human gait signatures of individuals who are carrying no visible load on their bodies (see Fig. 1) are distinct from those signatures of individuals with no load. The current challenge is to extend the National Institute of Standards and Technology Human ID at a Distance gait program started in 2001. That program concentrated on recognition capability of outdoor data with study of covariate factors such as lighting, footwear, apparel, surface conditions, and shadowing. CTISR’s program is investigating gait variance, not Human ID.

To investigate the current challenge problem involving carrying a concealed load under clothing, the capabilities of the CTISR Video Analysis and Content Extraction laboratory (Fig. 2), an outdoor gait track with a cross-over platform (Figs. 3 and 4), varying clothing types, vests weighing up to 10 kilograms (Fig. 5), and volunteer human subjects are all being used to create a rich database for signature development.

Design, construction, operation, and data collection and processing for Integration of a Sensor Package for Identification Radical Extremists (INSPIRE) are the responsibilities of two Southwestern Ohio Council for Higher Education students--Wright State University student Robert McGrellis and Cedarville student Jonathan Juhl. Two feature approaches are being pursued to address the challenge problem. The first is model-free analysis of silhouettes, being led by Maj. David M. Kaziska, Assistant Professor of Mathematics at AFIT's Graduate School of Engineering and Management. The second approach is model-based analysis, being led by Dr. Kendricks. Some preliminary, positive results have been obtained by the INSPIRE group, but as with any biometric investigation, considerable additional work is necessary to bring these approaches to fruition for field operations. The remainder of this article takes a closer look at their approaches to this area of research.

The initial results from Maj. Kaziska’s approach were encouraging. His method is a shape-based approach. This approach was first applied in a larger gait experiment. For each subject, training and test gait templates were computed; these templates were mean cycles over several cycles of gait. Persons in the test set are identified by finding the nearest match in the training set, using a distance metric between the shapes in the gait templates at comparable time points. This method was applied to preliminary data in the INSPIRE project by attempting to determine, for a single individual, whether or not that individual was wearing a pack on the torso. Video was shot of subjects with and without the pack, and training and test templates were computed. To determine whether or not a subject in the test set was wearing the pack, the metric is computed to the mean cycles from the same subject with and without the pack. In this initial study, 6 of 8 matches were correct.

Dr. Kimberly Kendricks will assist efforts to develop a set of characteristics in gait that are noticeable at a distance of seventy-five to one hundred feet to determine whether or not an individual may pose a threat. Dr. Kendricks’ mathematical research focus is in Groebner Basis Theory and its applications. For her dissertation, “Solving the Inverse Kinematic Robotics Problem: A Comparison Study of the Denavit-Hartenberg Matrix and Groebner Basis Theory,” she proved that using Groebner Basis theory, which is based upon a computational reduction algorithm, is a more advantageous approach to solving this problem. Since the human body moves like a robot, Dr. Kendricks has broadened her applications of this mathematical theory to Dr. Ron Tuttle’s gait analysis project, and through the inverse kinematics problem will apply Groebner Basis Theory to a mathematical model of arm-swing and leg-swing movement in the gait cycle and use hand and foot placement to help discover gait abnormalities. Thus far, Dr. Kendricks has developed a mathematical model describing arm-swing movement and verified it by discovering a Groebner basis that yields a one-to-one correspondence of solutions found through the forward and the inverse kinematic problems. Dr. Kendricks can now apply this mathematical model to the data currently being collected by Dr. Tuttle’s INSPIRE team.