Day/night underwater object detection from an airborne sensor using NOVAS (Non-acoustical Optical Vulnerability Assessment Software)

The majority of underwater object detections, from airborne platforms, are done with the unaided eye or a video camera during daytime or nighttime searches. An understanding of the optical environment in which detection is performed is, therefore, crucial in predicting the depth at which the underwater object can be detected day or night. The motivation behind NOVAS (Non-acoustical Optical Vulnerability Assessment Software) is to provide the Navy with a round-the-clock (24/7), PC-based assessment tool of the optical environment in which the warfighter operates. In order to achieve this goal, NOVAS has two types of predictions that are divided up according to the time of day: vertical visibility depth of an underwater target for the daytime, and bioluminescence signature of underwater man made objects as they navigate at different speeds in the water column at night. Both types of predictions are made by treating the optical environment as realistically as possible: a dynamic sea surface topography overlying a water column with depth-dependent optical properties. The topography of the sea surface depends on wind speed and far-off swell generation. The optical properties of the water column consist of absorption, scattering, and a mean scattering angle. The amount of bioluminescent activity generated by a moving source depends on both the biological organism concentration through which it is moving and its speed relative to the local currents. NOVAS also contains data maps that either display the location of the different experimental survey stations taken during a particular NAVOCEANO cruise, or are satellite generated. NOVAS models a low-light-level airborne camera with an intensified CCD. The loss of resolution in the intensification process is also modeled with the help of the Modulation Transfer Function. NOVAS predictions are displayed synoptically with regional maps and simulated images, as well as in one spot in the ocean, using in situ data. Images are simulated as if the viewer or system is looking down from an airborne platform hovering over an underwater object.