Hyperspectral machine vision as a tool for water stress severity assessment in soilless tomato crop

Abstract

Early detection of water deficit stress is essential for efficient crop management. In this study, hyperspectral machine vision was used as a non-contact technique for detecting changes in spectral reflectance of a soilless tomato crop grown under varying irrigation regimes. Four different irrigation treatments were imposed in tomato plants grown in slabs filled with perlite. The plants were grown in a growth chamber under controlled temperature and light conditions, and crop reflectance measurements were made using a hyperspectral camera to measure the radiation reflected by the crop from 400 nm to 1000 nm. The results showed that crop reflectance increased with increasing water deficit, and the detected reflectance increase was significant during the first day of irrigation was withheld. Based on the reflectance measurements, several crop indices were calculated and correlated with substrate volumetric water content and tomato leaf chlorophyll content. The results showed that when the modified red simple ratio tndex (mrSRI) and the modified red normalized vegetation index (mrNDVI) values increased by more than 2.5% and 23% respectively, the substrate volumetric water content decreased by more than 3%. In addition, when the Transformed Chlorophyll Absorption Reflectance Index (TCARI) value increased by about 16%, the leaf chlorophyll content decreased by about 3%. These results of the present study are promising for the development of a non-contact method for estimating plant water status in tomato crops grown under controlled environment. © 2017 IAgrE