期刊
JOURNAL OF OPTICS
卷 20, 期 4, 页码 -出版社
IOP PUBLISHING LTD
DOI: 10.1088/2040-8986/aab3b8
关键词
compressive sensing; computational imaging; hyperspectral; near-infrared
类别
资金
- Ministry of Science, Technology Space, Israel [3-18410, 3-13351]
- Kreitman Foundation
The acquisition of hyperspectral (HS) image datacubes with available 2D sensor arrays involves a time consuming scanning process. In the last decade, several compressive sensing (CS) techniques were proposed to reduce the HS acquisition time. In this paper, we present a method for near-infrared (NIR) HS imaging which relies on our rapid CS resonator spectroscopy technique. Within the framework of CS, and by using a modified Fabry-Perot resonator, a sequence of spectrally modulated images is used to recover NIR HS datacubes. Owing to the innovative CS design, we demonstrate the ability to reconstruct NIR HS images with hundreds of spectral bands from an order of magnitude fewer measurements, i.e. with a compression ratio of about 10:1. This high compression ratio, together with the high optical throughput of the system, facilitates fast acquisition of large HS datacubes.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据