FPGA Acceleration of a Supervised Learning Method for Hyperspectral Image Classification

Abstract

Hyperspectral image classification is one of the most important techniques for analyzing hyperspectral image that have hundreds of spectrum luminance values. For this classification, supervised learning methods are widely used, but in general, they have a trade-off between their accuracy and computational complexity. In this paper, we propose an FPGA implementation of hyperspectral image classification based on a composite kernel method. Because of the large size of hyperspectral images, the data mapping becomes the most critical issue for achieving higher processing speed. Two data mapping approaches are discussed, and one of them that is most suitable for our target images is implemented on an FPGA. Its processing speed for 145×145 pixel images is fast enough for real-time processing, and its accuracy is comparable with other classification algorithms.