This paper examines imaging performance bounds for undersea electro-optic identification (EOID) sensors that use
pulsed laser line scanners to form serial images, typically utilizing one laser pulse for each formed image element. The
experimental results presented include the use of two distinct imaging geometries; firstly where the laser source and
single element optical detector are nearly co-aligned (near monostatic) and secondly where the laser source is deployed
on a separate platform positioned closer to the target (bistatic) to minimize source-to-target beam spread, volumetric
scatter and attenuation, with the detector being positioned much further from the target. The former system uses
synchronous scanning in order to significantly limit the required instantaneous angular acceptance function of the
detector and has the desired intention of acquiring only ballistic photons that have directly interacted with the target
element and the undesirable property of acquiring snake photon contributions that indirectly arrive into the detector
aperture via multiple forward scattering over the two-way propagation path. The latter system utilizes a staring detector
with a much wider angular acceptance function, the objective being to deliver maximum photon density to each target
element and acquire diffuse, snake and ballistic photon contributions in order to maximize the signal.
The objective of this work was to experimentally investigate pulse-to-pulse detection statistics for both imaging
geometries in carefully controlled particle suspensions, with and without artificially generated random uncharacterized
scattering inhomogeneities to assess potential image performance in realistic conditions where large biological and
mineral particles, aggregates, thin biological scattering layers and turbulence will exist. More specifically, the study
investigates received pulse energy variance in clear filtered water, as well as various well-characterized particle
suspensions with and without an artificial thin random scattering layer. Efforts were made to keep device noise constant
in order to assess the impact of the environment on extrapolated image quality.
|