Abstract:In order to solve the problem that drought stress and salt stress are difficult to distinguish in the early stage when plants are under stress, a method was proposed to identify drought stress and salt stress based on light-induced plant electrical signals. The illumination/darkness cycle stimulation was used to obtain the surface potentials of wheat seedlings under normal conditions, as well as under isotonic drought and salt stress. One-versus-one support vector machine (OVOSVM) was used to classify the obtained plant electrical signals. The results of 3-fold cross validation showed that the two-class classification of wheat seedling leaf surface potential under normal and drought stress had an accuracy of 100%, and the two-class classification of leaf surface potential under normal and salt stress had an accuracy of 94.44%. The accuracy of three-class classifications reached 96.30%. Under the conditions of isosmotic drought stress and salt stress, the classification accuracy of wheat seedling leaf surface potential was 100%. The results showed that plant electrical signals can be used as a method to identify adversity stress, and can accurately identify early drought stress and salt stress under isotonic conditions.