Restoration of Missing Lines in Grip Patterns for Biometrics Authentication on a Smart Gun

The Secure Grip project aims to develop a grip-pattern recognition system, as part of a smart gun. Its target users are the police officers. The current authentication algorithm is based on a likelihood-ratio classifier. The grip pattern is acquired by sensors on the grip of the gun. Since in practice various factors can result in missing lines in a grip pattern, restoration of these missing lines will be useful and practical. We present a restoration algorithm based on null-space error minimization. The simulation results of the restoration and authentication experiments show that this restoration algorithm effectively restores grip patterns, and is, therefore, capable of improving the system’s authentication performance when missing lines are present.