Detection Of Crop Diseases And Insect Pests Based On Convolutional Neural Network

This study is based on a convolutional neural network model built with deep learning technology, and compares the mainstream target detection algorithms ResNet v2, Mask R-CNN, and YOLOv7 on the same tomato pest and disease data set. Select the algorithm with the best accuracy, average accuracy, and F1 score based on the model evaluation indicators. Experimental results show that the average accuracy (mAP) of ResNet v2 is 92.82%, Mask R-CNN is 94.90%, YOLOv7 is 95.17%, and IP YOLOv7 tomato pest detection is 97.35%, surpassing ResNet v2, Mask R-CNN and other networks. This study provides certain reference and inspiration for the development of tomato pest and disease detection technology, promotes the protection of ecosystem balance, ensures safe crop production, improves the level of agricultural intelligence, and contributes to achieving sustainable agricultural development and ensuring food security.