A Performance Driven Framework for Real-time Exploitation of OPIR Imagery

As the number of OPIR sensors and the corresponding resolution and frame rates increase, the cost of transmitting the raw data stream to the ground station and moving this data between analyst workstations is rapidly increasing. Because of this, the incorporation of real-time processing both at the sensor and the ground station to characterize regions of interest is becoming a crucial component of the analyst workflow. In this proposed effort, we will develop a novel real time algorithm for identification and separation of multiple closely-spaced, dim targets in short-wave infrared data acquired by space-based sensors. Our approach leverages our history in OPIR algorithm research and optimization, extending the current state of the art utilizing proven techniques for clutter suppression in conjunction with novel super-resolution and deconvolution algorithms. Further, by modeling the performance of the algorithms as a function of key operating conditions (SNR, frame rate, resolution, pointing stability), we will develop a framework for analyzing the impact of on-board vs. centralized ground station vs. analyst workstation based implementations of the algorithms. As part of our commercialization strategy, we will integrate our algorithms within the JAGWire/HYDRA environment.