|Authors: ||N. Katsoulas, A. Elvanidi, K.P. Ferentinos, C. Kittas, T. Bartzanas|
|Keywords: ||camera, machine vision, noise elimination, irrigation scheduling, NDVI|
Although much progress has been made on optimizing plant water supply, only a limited number of methods use plant-based physiological indicators to detect plant water stress and adapt irrigation scheduling accordingly.
In addition, even fewer indicators can be estimated remotely without contact and effect on plant development.
Hyperspectral imaging could be an accurate technique to detect plant water status, taking into account crop characteristics.
In this work, a methodology of hyperspectral imaging calibration and acquisition is presented.
The technique uses the crop reflectance characteristics from 400 to 1000 nm and incorporates the appropriate radiometric and geometric corrections.
It was confirmed that sensor's dark current noise is proportional to exposure time and frame rate values, while CCD silicon detector is wavelength-dependent.
The basic statistical parameters of mean and standard deviation values were used to estimate spatial and spectral correlation of each band on the extracted areas/pixels of interest.
Several statistical techniques were used for the selection of optimal features that would lead to the development of appropriate plant water stress indices that could be used for incipient water stress detection in optimal irrigation scheduling systems.
The images were clearer when the exposure time was 130 ms and the speed of the scanner was set at 0.16 mm s-1 with a frame rate of 500 Hz.
NDVI, rNDVI and mrNDVI indices proved to be independent of light signal variation.
Download Adobe Acrobat Reader (free software to read PDF files)
URL www.actahort.org Hosted by KU Leuven