Currently, the plant physiological information detection technology and diagnostic systems has gradually become a hot topic in the research of facilities crop disease warning. By collecting the relevant data of changes in growth conditions of cucumber by Junior-PAM of CI-340 portable photosynthesis system, the experiment studied the correlation between leaf chlorophyll content and Photosynthetic rate, nutritional status and we like to search for the relationship between the changes in disease infection and chlorophyll content in the growth process of leaf. The experiment used different modeling methods to establish the relational model between fluorescence data such as FO ', FM', FV / FM, FO, FM and SPAD, and then make predictions. The results show that the prediction of harmful levels of disease employing neural network model is the best with its MAE being 0.0025 and the accuracy up to 100%.