Early change detection of plant leaf water content using polarization imaging

Water content is crucial to plant growth. Water shortage can cause a decrease in photosynthesis and respiration efficiency during growth, directly leading to the yield of plants and crops. An accurate identification can greatly contribute to the water change of plant leaves in the early stage of water shortage and intervene rapidly. It can greatly reduce the loss of plants and crops due to water shortage. However, there is little change in the water content and no outstanding change in the plant morphology and spectral characteristics during the early stage. It is a high demand for the early identification of plant water content change at present. Two measurements can be widely used in the plant water content. One is the direct measurement (such as drying, and distillation), where the measurement accuracy is higher, generally as the standard of other methods. Another is the indirect measurement using, such as the plant stem diameter, canopy temperature, conductivity, microwave, spectroscopy, and images. In this study, a set of polarized light imaging systems was built to add the light source, and polarizer in the image detection. The polarized light imaging detection was then carried out on the holly leaves with different water content. The polarization of light that carried the multidimensional vector information was used as the information carrier to transmit more abundant information. The multi-dimensional vector that carried in the light polarization information was used to detect the different water content of holly leaves, in order to realize the early identification of water content change. The main conclusions of this study were as follows. 1) The m₁₃ and m₃₂ matrix elements in the Mueller matrix image initially reflected the change of water content of holly leaves, but there was no outstanding difference in the early stage of water content change. 2) The depolarization index image clearly showed the change in the water content of holly leaves at each stage. Specifically, the depolarization index increased with the continuous decrease of the water content of holly leaves. This was because the cells inside the leaves were dehydrated gradually, when the water content of the leaves decreased gradually, leading to the increase of multiple scattering. Thus, the overall depolarization index of leaves increased as well. 3) The mapping relationship was obtained between the water content and depolarization index as after fitting. The determination coefficient of the fitting curve was 0.95. Then, different parts of holly leaves were selected to verify the prediction model. The maximum relative error between the predicted water content and the actual was 4.90%. The polarized light imaging was achieved in the high contrast, strong background noise suppression, and more information, compared with the current commonly used image detection. Furthermore, the cells inside the leaves gradually perished in the process of continuous water loss of holly leaves, leading to the increase in the multiple scattering of light into the leaves and the increase of depolarization index. Therefore, the depolarization index can be used as an important parameter to identify the early changes in the water content of plants and crops. This finding can provide new insight for the early identification of the water changes in plants and crops. A theoretical basis can also be proved for the rapid nondestructive testing of physiological states, such as salt, acid, and fertilizer stress. Therefore, the integration degree of the system can be optimized to improve the portability of the system in the future. The wave plate can also be replaced with the liquid crystal phase retarder to improve the speed and accuracy of image acquisition. The sample database can be established to constantly optimize the improved model for better accuracy of prediction.