Single-shot Hyperspectral Fluorescent Imaging

2017 Fiscal Year Final Research Report

• Project/Area Number: 16K16095
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Perceptual information processing
• Research Institution: National Institute of Informatics
• Principal Investigator: ZHENG YINQIANG 国立情報学研究所, コンテンツ科学研究系, 助教 (30756896)
• Research Collaborator: SATO Imari ; MEGURO Misaki ; ASANO Yuta ; JIA Yan ; GU Lin ; SATO Yoichi
• Project Period (FY): 2016-04-01 – 2018-03-31
• Keywords: スペクトルイメージング / ハイパースペクトルカメラ / マルチスペクトル / スペクトル再構成

Outline of Final Research Achievements

There are many objects and substances around us, like dyes and plants, which contain both the reflective and fluorescent components. To separate the reflective component from the fluorescent component is helpful in examining the intrinsic status of objects. Therefore, effective methods for this separation task are needed. Existing methods for fluorescence and reflectance separation needs multiple shots under different illuminations, thus are inapplicable to dynamic scenes. This research first uses a single hyperspectral image, and designs the optimal illumination spectrum for this separation task. The fact that existing hyperspectral cameras are still not very fast in data capture leads us to explore the possibility of reconstructing hyperspectral images from RGB images. On the application aspect, a simple separation method of weak fluorescence in the presence of strong environmental illumination has been developed and successfully used for freshness examination of meat and cheese.

Free Research Field

コンピュータビジョン、コンピュテーショナルフォトグラフィ