Turbulence degradation and mitigation performance for handheld weapon ID

Abstract

Atmospheric turbulence can severely limit the range performance of state-of-the-art large aperture imaging sensor systems, specifically those intended for long range ground to ground target identification. Simple and cost-effective mitigation solutions which operate in real-time are desired. Software-based post-processing techniques are attractive as they lend themselves to easy implementation and integration into the back-end of existing sensor systems. Recently, various post-processing algorithms to mitigate turbulence have been developed and implemented in real-time hardware. To determine their utility in Army-relevant tactical scenarios, an assessment of the impact of the post processing on observer performance is required. In this paper, we test a set of representative turbulence mitigation algorithms on field collected data of human targets carrying various handheld objects in varying turbulence conditions. We use a controlled human perception test to assess handheld weapon identification performance before and after turbulence mitigation post-processing. In addition, novel image analysis tools are implemented to estimate turbulence strength from the scene. Results of this assessment will lead to recommendations on cost-effective turbulence mitigation strategies suitable for future sensor systems.