A Low-Complexity Hyperspectral Anomaly Detection Algorithm and Its FPGA Implementation

On-board real-time anomaly detection has always been a challenging task in hyperspectral imaging analysis as it requires low computational complexity. Most of the existing algorithms inevitably trade off intensive complexity for high accuracy. This article presents fast spectral-spatial algorithm with images (HSIs) using morphological reconstruction and simplified guided filter (Fast-MGD). Since simple filtering techniques are applied, is therefore feasible to achieve field programmable gate array (FPGA)-based hardware implementation. More precisely, an effective deeply pipelined acceleration scheme developed adopting high-level synthesis support HSIs that acquired over different scenes sizes spectral bands. Experimental results show strong advantages proposed FPGA-based Fast-MGD processing speed resource consumption, while accuracy remained. Its applicability on-board demonstrated verifie.