FPGA IMPLEMENTATION OF A MAXIMUM VOLUME ALGORITHM FOR ENDMEMBER EXTRACTION FROM HYPERSPECTRAL IMAGERY

Endmember extraction is a very important technique in hyperspectral unmixing. Due to the fact that more applications require real or near real-time processing capabilities, high performance computing based on field programmable gate array (FPGA) for endmember extraction has received considerable interest in recent years. In this paper, we propose a real-time implementation of maximum volume algorithms (RT-MaxV) using a Kintex-7 FPGA. The proposed RT-MaxV does not need dimensionality reduction like other real-time fast simplex growing algorithms (RT-FSGA). Our experimental results, obtained using the AVIRIS Cuprite data set, indicate that the proposed algorithm has better accuracy and performance than RT-FSGA, and its FPGA implementation achieves real-time processing capability in the considered problem.