Real-Time Image Restoration for Generation 3 Forward Looking Infrared Systems

Integrated Sensors, Incorporated (ISI) proposes to evaluate a variety of image restoration algorithms for a compact, real-time, high-resolution FLIR implementation. The target hardware solution is a field-programmable gate array (FPGA) multiprocessorarchitecture. These provide the speed of a hardware implementation, while offering the reprogramming flexibility of a general-purpose processor.Algorithms such as adaptive Wiener filters and recursive maximum likelihood / entropy optimizations are amongthose under consideration. All of these require precise models of the system MTF (Modulation Transfer Function). The Wiener approach is nonrecursive and would be the simplest to implement, while the recursive approaches tend to perform better undernongaussian conditions. A major challenge is to improve the convergence time of the recursive approaches. Innovation is also required to structure these algorithms for an efficient, real-time solution. The algorithms will be evaluated using restorationperformance and computational requirements as key discriminants. The performance will be evaluated subjectively on real images and quantified using minimum resolvable temperature difference (MRTD) measurements on simulated data. A hardware solution willbe customized to support the most promising algorithm.ISI will include this technology as part of its ImageExpressT commercial motion analysis system, to enhance measurement accuracy.Commercial applications include the enhancement of ISI's ImageExpressTproduct line and marketing the technology to vendors of high resolution FLIR and consumer video systems.