Research on recognition method of leaf diseases of woody fruit plants based on transfer learning

Fruit leaf diseases have a significant impact on the later development and maturity of fruits, so rapid and accurate identification of fruit leaf diseases plays an important role in the development of fruit production. In this paper, the leaf disease data set of 6 kinds of fruits is divided into 25 categories according to the species—the type of the disease—the severity, and we propose an improved model based on ResNet101 to identify woody fruit plant leaf diseases, in which a global average pooling layer is used to reduce model training parameters, layer normalization, dropout and L2 regularization are used to prevent model overfitting, SENet attention mechanism is used to improve the model's ability to extract features. At the same time, transfer learning is used to reduce training time and training parameters. Experimental results show that the overall accuracy of woody fruit plant leaf recognition based on this model can reach 85.90%. Compared with the classic ResNet network, the accuracy is increased by 1.20%, and the model parameters are reduced by 98.14%. Therefore, the model proposed in this paper provides a better solution for the identification of leaf diseases of woody fruit plants and has a higher accuracy rate.