Highly sensitive image-derived indices of water-stressed plants using hyperspectral imaging in SWIR and histogram analysis

David M. Kim, Hairong Zhang, Haiying Zhou, Tommy Du, Qian Wu, Todd C. Mockler, Mikhail Y. Berezin

Research output: Contribution to journalArticle

24 Scopus citations

Abstract

The optical signature of leaves is an important monitoring and predictive parameter for a variety of biotic and abiotic stresses, including drought. Such signatures derived from spectroscopic measurements provide vegetation indices - a quantitative method for assessing plant health. However, the commonly used metrics suffer from low sensitivity. Relatively small changes in water content in moderately stressed plants demand high-contrast imaging to distinguish affected plants. We present a new approach in deriving sensitive indices using hyperspectral imaging in a short-wave infrared range from 800 nm to 1600 nm. Our method, based on high spectral resolution (1.56 nm) instrumentation and image processing algorithms (quantitative histogram analysis), enables us to distinguish a moderate water stress equivalent of 20% relative water content (RWC). The identified image-derived indices 15XX nm/14XX nm (i.e. 1529 nm/1416 nm) were superior to common vegetation indices, such as WBI, MSI, and NDWI, with significantly better sensitivity, enabling early diagnostics of plant health.

Original languageEnglish
Article number15919
JournalScientific reports
Volume5
DOIs
StatePublished - Nov 4 2015

Fingerprint Dive into the research topics of 'Highly sensitive image-derived indices of water-stressed plants using hyperspectral imaging in SWIR and histogram analysis'. Together they form a unique fingerprint.

  • Cite this