Synthetic Aperture LADAR Automatic Target Recognizer Design and Performance Prediction via Geometric Properties of Targets (Preprint)
Abstract
Synthetic Aperture LADAR (SAL) has several phenomenology differences from Synthetic Aperture RADAR (SAR) making it a promising candidate for automatic target recognition (ATR) purposes. The diffuse nature of SAL results in more pixels on target. Optical wavelengths offers centimeter class resolution with an aperture baseline that is 10,000 times smaller than an SAR baseline. While diffuse scattering and optical wavelengths have several advantages, there are also a number of challenges. The diffuse nature of SAL leads to a more pronounced speckle effect than in the SAR case. Optical wavelengths are more susceptible to atmospheric noise, leading to distortions in formed imagery. While these advantages and disadvantages are studied and understood in theory, they have yet to be put into practice. This dissertation aims to quantify the impact switching from specular SAR to diffuse SAL has on algorithm design. In addition, a methodology for performance prediction and template generation is proposed given the geometric and physical properties of CAD models. This methodology does not rely on forming images, and alleviates the computational burden of generating multiple speckle fields and redundant ray-tracing.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 10, 2022
- Accession Number
- AD1159402
Entities
People
- Jacob W. Ross
Organizations
- Air Force Research Laboratory