Detection of small targets in ocean wave clutter using panchromatic time series imagery

Visual detection of small objects on or near the ocean surface from an aircraft is important for search and rescue, and detection of stealthy boats and other militarily important targets. This is made difficult by a combination of the rapidly moving sensor platform and the typically high level of modulations in scattered light at the surface. To achieve reasonable search rates, the observer must scan through modest grazing angles (i.e., look towards the horizon) where the ocean clutter that limits detection performance is varying light from wave facets and white caps. This paper describes a technique for significantly improving the signal-to-clutter ratio. A time series of images is mapped to a common reference frame at the level of the mean ocean surface, separating time and space variations, and the temporal dwell is utilized. For moored targets, the signal may be simply integrated up in time, while for drifting and powered targets, it must be integrated along the appropriate target velocity vector. Specific examples are provided for navigation buoys, small surface floats and slowly moving boats, most of which are difficult if not impossible to find in any single image frame. Achieved gain using panchromatic images is shown to approach 20 dB in cases where the velocity vector is known or can be calculated from the data.