基于分焦平面偏振成像的冬青卫矛含水率快速无损检测.

Water has engaged in a series of activities during plant growth. Therefore, the moisture content is one of the most important indicators of plant growth. Water deficiency can lead to a decrease in the stomatal degree on the leaf surface, where the carbon dioxide can flow into the plant leaf. The resulting photosynthetic efficiency can greatly reduce the appearance, nutrition, and total biomass of forestry plants. However, it is still lacking in the rapid detection of moisture content in plants. In this study, a polarization imaging system was proposed using a split-focal plane polarization camera. Polarization images were then analyzed to establish the dependence between the linear polarization and moisture content, in order to reveal the interaction between polarized light and water in plants. The rapid and nondestructive detection was realized for the moisture content of wintergreen leaves. The specific research was as follows. 1) The holly leaves were first collected and dried in a gradient manner. The moisture content was obtained after weighing; 2) A sub-focal plane polarization system was developed to detect the polarization of the gradient-dried holly leaves; 3) A systematic investigation was carried out to clarify the influence of growth states on the moisture content of holly leaves under various months, different parts, as well as sunny or rainy weather conditions. 40 leaves were collected and then divided into 8 groups, with 5 leaves for the control in each group; 4) The holly leaves were detected for the polarization. Stokes parameters were obtained to convert into the linear polarization degrees with clear physical significance and imaging. It was found that the linear polarization degree increased gradually, as the moisture content decreased. There was a strong negative correlation between the leaf moisture content and the linear polarization degree. 5) Rapid detection of moisture content was achieved by the quantitative relationship between moisture content and linear polarization degree after fitting. The degree of linear polarization was only used to detect the moisture content of the target leaf. 6) Other holly leaves were collected and then substituted into the fitting model, in order to verify the prediction of the model. The results show that the moisture content of holly spear leaves was represented by the linear polarization degree in the three growth conditions (months, weather, and parts). Furthermore, the linear polarization degree increased gradually with the decrease in the gradient of moisture content. A prediction model was obtained using the large number of plant leaves. The correlation coefficient reached 0.85, and the MSE was only 0.45%. Verification through the leaf moisture detection system showed that the average relative error of the moisture detection system was 9.50%, and the RMSE was 4.53%. The true moisture content of the leaf was characterized by linear polarization degree. The findings can also provide optical applications in plant health detection. [ABSTRACT FROM AUTHOR]