Automatic Real-time Computer Aided IDentification (ARCAID)

The AQS-20(A) towed minehunting system incorporates an electro-optic imaging lidar sensor that provides high fidelity 3D imagery crucial for identification of mines. While a real-time algorithm to automatically detect, classify, and identify targets fromthe image outputs of the EO system is entirely feasible, the current AQS-20(A) system support only a rudimentary Automatic Target Detection (ATD) algorithm that has limited capability to differentiate mine-like objects (MLOs) from clutter.Aret¿ proposes development of an Automatic Real-time Computer Aided IDentification (ARCAID) system that would replace the current ATD and greatly expedite target identification and false alarm rejection during AQS-20(A) missions. The ARCAID system willgreatly reduce the amount of post-mission analysis required to positively identify mines and discriminate them from mine-like clutter because the real-time algorithm will provide a high level of identification and clutter discrimination on the fly.Additionally, the ARCAID system will incorporate a 3D data visualization tool for man in-the-loop verification of mine identifications. The tool will enable the operator to optimally exploit the 3D data captured by the system by interactively comparingmultiple 3D perspective views of mine-like objects (MLOs) identified by the algorithm with identically scaled models of the suspected mines. Aret¿ sees the ARCAID system as a necessary and integral part of the AQS-20(A) EOID system. Because Aret¿ already has experience developing a baseline algorithm for the ARCAID system, Aret¿ believes that it would be a grave mistake to omit such acapability in the fleet-deployment of the AQS-20(A) system. The ARCAID system will greatly enhance the efficiency and accuracy of the AQS-20 mine-hunting mission, which will provide a huge asset to national defense and to the efficacy of our naval warfare.