Abstract
Due to increased security concerns, the commitment to monitor and maintain security in the maritime environment is
increasingly a priority. A country's coast is the most vulnerable area for the incursion of illegal immigrants, terrorists
and contraband. This work illustrates the ability of a low-cost, light-weight, multi-spectral, multi-channel imaging
system to handle the environment and see under difficult marine conditions. The system and its implemented
detecting and tracking technologies should be organic to the maritime homeland security community for search and
rescue, fisheries, defense, and law enforcement. It is tailored for airborne and ship based platforms to detect, track
and monitor suspected objects (such as semi-submerged targets like marine mammals, vessels in distress, and drug
smugglers). In this system, automated detection and tracking technology is used to detect, classify and localize
potential threats or objects of interest within the imagery provided by the multi-spectral system. These algorithms
process the sensor data in real time, thereby providing immediate feedback when features of interest have been
detected. A supervised detection system based on Haar features and Cascade Classifiers is presented and results are
provided on real data. The system is shown to be extendable and reusable for a variety of different applications.
