A 3D Hyperspectral Imagery Simulation Test Bed for the Littoral Zone

Period of Performance: 04/18/2000 - 10/18/2000

$100K

Phase 1 SBIR

Recipient Firm

Spectral Sciences, Inc.
4 Fourth Avenue Array
Burlington, MA 01803
Principal Investigator

Abstract

Remote sensing of the ocean-land interface, the littoral zone, is a valuable source of information supporting military and commercial applications, such as Littoral Warfare Data (LWD) for amphibious assaults and fish habitat monitoring in coastal waters. The quality of the retrieved information depends critically on the accuracy of the atmospheric correction and the surface/material identification and feature/object extraction algorithms. This proposal addresses the need for an accurate, robust, and efficient means for algorithm development and validation for littoral zone applications. We propose a high fidelity hyperspectral image simulation test bed for VIS-NIR (400-2500nm) sensors tailored to the littoral zone, dubbed SHALLOWS (Synthetic Hypercubes for Amphibious Littoral Landings with Objects under Water Simulation). The Phase I objectives include: (1) extending a 3D Monte Carlo land surface simulation model into the littoral zone, (2) validating against 1D models, (3) a feasibility demonstration simulation for a littoral zone scenario including submerged objects, and (4) developing a littoral zone atmospheric correction approach. In Phase II, advanced versions of the simulation and atmospheric correction algorithms will be developed, coded into versatile software packages, and applied to analysis of commercial satellite littoral zone imagery with an emphasis on LWD information extraction. We envision two commercial software products resulting from this effort, one for simulation of airborne and satellite-based panchromatic, multi and hyperspectral VIS-NIR imagery of the littoral zone, and the other for littoral zone atmospheric correction and surface, feature, and object extraction and identification. Both products will address military and commercial market needs for user-friendly analyst-oriented and fully automated software packages that facilitate the analysis of VIS-NIR imagery. The range of applications is extensive; a few examples include mission planning for amphibious assaults, battlefield surveillance, target detection and identification, technical intelligence, pollutant monitoring of coastal regions and inland waters, monitoring of fish habitats, and damage assessment from coastal storms.