Water stress detection in Sunagoke moss ( Rhacomitrium canescens) using combined thermal infrared and visible light imaging techniques

Abstract

This study is a contribution to an ongoing research to develop a thermal-reflectance imaging system (TRIS) for detecting water stress in Sunagoke moss (Rhacomitrium canescens). The correlation between visual and thermal water stress symptoms in a sample of Sunagoke moss as a function of time and water content was evaluated. Visible light imaging was used to report changes in the surface structure and reflectance. Infrared (IR) thermal imaging was used to monitor transpiration patterns and the crop water stress index (CWSI) was used to quantify those changes. Grey-level concurrence matrix (GLCM) texture features were used to quantify changes in the sample surface structure while the RGB colour ratios were used to detect changes in its reflectance. The sample exhibited water-related stress at water contents below 1.5 g g−1 and above 3.0 g g−1 (grams of water per gram of dry sample). The maximum possible total uncertainty of IR temperatures for the sample was ±0.50 °C at 17 °C, ambient temperature. The maximum uncertainty of its visible light data was ±7.65 grey level attributed to the red bandwidth. The results showed a clear correlation between the water status of the sample and its CWSI, GLCM texture and RGB colour ratios. These results demonstrate the possibility of detecting both drought and flood water stress in Sunagoke moss by combining thermal and visible light imaging. Although Sunagoke moss was used in this study, this method is novel and could be extended to both biotic and abiotic stress detection in other plants.