Automatic extraction of canopy and artificial reference temperatures for determination of crop water stress indices by using thermal imaging technique and a fuzzy-based image-processing algorithm

Thermal stress indicators are one of the most accurate indices for sensing plant water status that can be remotely measured by the means of infrared thermography. In addition to the canopy temperature, these indices need to access the wet and dry reference temperatures which refer to the temperatures of the canopy at well-watered and fully stressed conditions, respectively. The main goal of this study is to measure the canopy as well as reference temperatures automatically by the means of a single thermal image, captured from an olive tree. The temperatures of artificial reference surfaces were extracted by the means of an object detection method based on the edge detection and morphological processes. The temperatures of sunlit and shaded canopy portions were also detected, using a Fuzzy C-means clustering of thermal images with the wet and dry reference temperatures as thresholds. The algorithm was successfully detected the references in 90% of the images and the automatic extracted canopy temperatures were significantly correlated with the manual ones.