Radiative transfer modeling and analysis of spatially variant and coherent illumination for undersea object detection

Increasing the optical range of target detection and recognition continues to be an area of great interest in the ocean environment. Light attenuation limits radiative and information transfer for image formation in water. These limitations are difficult to surmount in conventional underwater imaging system design. Methods for the formation of images in scattering media generally rely upon temporal or spatial methodologies. Some interesting designs have been developed in an attempt to circumvent or overcome the scattering problem. This document describes a variation of the spatial interferometric technique that relies upon projected spatial gratings with subsequent detection against a coherent return signal for the purpose of noise reduction and image enhancement. A model is developed that simulates the projected structured illumination through turbid water to a target and its return to a detector. The model shows an unstructured backscatter superimposed on a structured return signal. The model can predict the effect on received signal to noise of variations in the projected spatial frequency and turbidity. The model has been extended to predict what a camera would actually see so that various noise reduction schemes can be modeled. Finally, some water tank tests are presented validating original hypothesis and model predictions. The method is advantageous in not requiring temporal synchronization between reference and signal beams and may use a continuous illumination source. Spatial coherency of the beam allows detection of the direct return, while scattered light appears as a noncoherent noise term. Both model and illumination method should prove to be valuable tools in ocean research.