MFDS-m Red Edge Position Detection Algorithm for Discrimination Between Healthy and Unhealthy Vegetable Plants

Abstract

Spectral Reflectance of crop shows very distinguished sensitivity in spectral regions according to biophysical and biochemical parameters. Red Edge Position is the inflection point in the red edge region of electromagnetic spectrum which is between 680–780 nm. This is sensitive indicator of crop health. Red Edge Position is used to discriminate between healthy and unhealthy plants. Analytical Spectral Devices (ASD) Fieldspec spectroradiometer instrument having spectral range from 350 nm to 2500 nm, was used to collect lab spectra of vegetable plants. An algorithm is proposed based on a Maximum First Derivative Spread – mean and its reflectance magnitude in Red Edge Region. Maximum First Derivative Spread-mean (MFDS-m) algorithm is proposed to detect Red Edge Position which will be further used to discriminate healthy and unhealthy plants. Results are compared with Four Point Linear Interpolation, Extrapolation and Maximum First Derivative Techniques.