Optimizing CNN-Based Hyperspectral Image Classification on FPGAs

Abstract

Hyperspectral image (HSI) classification has been widely adopted in remote sensing imagery analysis applications which require high classification accuracy and real-time processing speed. Convolutional neural networks (CNNs)-based methods have been proven to achieve state-of-the-art accuracy in classifying HSIs. However, CNN models are often too computationally intensive to achieve real-time response due to the high dimensional nature of HSI, compared to traditional methods such as Support Vector Machines (SVMs). Besides, previous CNN models used in HSI are not specially designed for efficient implementation on embedded devices such as FPGAs. This paper proposes a novel CNN-based algorithm for HSI classification which takes into account hardware efficiency and thus is more hardware friendly compared to prior CNN models. An optimized and customized architecture which maps the proposed algorithm on FPGA is then proposed to support real-time on-board classification with low power consumption. Implementation results show that our proposed accelerator on a Xilinx Zynq 706 FPGA board achieves more than 70$\times$ faster than an Intel 8-core Xeon CPU and 3$\times$ faster than an NVIDIA GeForce 1080 GPU. Compared to previous SVM-based FPGA accelerators, we achieve comparable processing speed but provide a much higher classification accuracy.

The first two authors contributed equally.