Real-time Atmospheric Disturbance Compensation Using Hardware-Accelerated Speckle Imaging

Atmospheric disturbances are a major performance-limiting factor in long-range optical systems. In particular, for the Airborne Laser (ABL) the ability of distinguish targets from a long distance is crucial for mission success. Despite the progress in optics and sensor technology, blurring in long-range imaging caused by atmospheric movements and density changes will remain an issue. Digital signal processing techniques can be used to compensate for these atmospheric effects, using algorithms like the bispectrum speckle method developed by researchers at Lawrence Livermore National Laboratories. Unfortunately, these algorithms are computationally intensive and require several seconds to process a single frame in high-end workstations, making them unsuitable for real-time video surveillance applications. We propose the development of a custom hardware processor, based on FPGA technology, which is able to implement the speckle algorithm two orders of magnitude faster than current PCs, thereby enabling real-time video feed processing. To this end, we plan to collaborate with the creators of the speckle algorithm at LLNL. Furthermore, FPGAs are uniquely suited for airborne platforms given their footprint and power consumption. For this project, we will develop a real-time atmospheric compensation solver that occupies less than 120 cubic inches and consumes less than 25 W of power.