CLASSIFICATION OF STRESS-INDUCED SPATIAL VARIABILITY IN NEAR-INFRARED CANOPY REFLECTANCE.
in HortScienceA digital video camera (Panasonic Industrial Co., Secaucus, NJ) with a 1.7 cm charged coupled device detector (574 (h) × 499 (v) pixel elements) was modified with a custom made FRF-700 band pass filter to visualize canopy reflectance in the near-infrared (NIR) from 700 to 1100 nm. Images of canopy reflectance under a range of environmental stresses were obtained from peach and apple trees under greenhouse and field conditions. Individual video frames were digitized with Image Capture and Analysis System (Agri Imaging Systems, Inc., Fayetteville, AR). Image contrast was increased with digital equalization and filtering before classification into one of five stress levels. There was a high correlation (r2 > 0.8) between leaf stress and canopy reflectance in both apple and peach at distances < 5 meters. Spatial variability in stress-induced NIR reflectance could be detected and classified at vertical distances from 150 to 500 M. Analysis of vertical imagery revealed sections of the orchard which were most susceptible to environmental stress.
