Walnut pest recognition system and detection method based on convolutional neural network

The invention discloses a walnut insect pest recognition and detection method based on a convolutional neural network; the method comprises the following steps: 1, data collection: obtaining corresponding insect pest pictures through employing a web crawler method because the number of pictures of diseases and pests is small, and screening out different pictures as a test set after the repeated pictures are preliminarily deleted, wherein since the difference between the number of photos of different insect pests is huge, the training effect of the convolutional network is poor, and the performance is reduced, a data enhancement method is adopted to carry out data enhancement on the insect pests with small data volume, and the difference between the data volumes of species is small. According to the method, the generalization ability of the model is improved, the operation time is shortened, and the storage space is reduced by relying on the convolutional neural network with better performance in machine learn