Automatic Identification of Peanut-Leaf Diseases Based on Stack Ensemble

Peanut is an important food crop, and diseases of its leaves can directly reduce yield quality. In order to solve the problem automatic identification peanut-leaf diseases, this paper uses a traditional machine-learning method ensemble output deep learning model identify peanut leaves. The included healthy leaves, rust disease on single leaf, leaf-spot scorch both leaf. Three types data-augmentation methods were used: image flipping, rotation, scaling. experiment, deep-learning had higher accuracy than methods. Moreover, achieved better performance when using data augmentation stacking ensemble. After by logistic regression, residual network with 50 layers (ResNet50) was as high 97.59%, F1 score dense convolutional 121 (DenseNet121) 90.50. used in experiment greatest improvement after regression Deep-learning networks deeper like ResNet50 DenseNet121 performed experiment. This study provide reference for diseases.