One of the most significant elements that pose a significant risk to agricultural productivity is the presence of agricultural diseases and insect pests. Finding and naming pests as soon as they appear is one of the most efficient ways to cut down on the financial damage they do. In this study, a convolutional neural network was utilized to automatically detect illnesses that might affect crops. This data set was obtained from the public data set that was provided for the AI Challenger Competition in 2018. It contains 27 photos of diseases affecting 10 different crops. Training is carried out with the help of the Inception-ResNet-v2 model in this article. The direct edge in the cross-layer and the multi-layer convolution in the residual network unit of the model. Following the completion of the combined convolution process, it is triggered by the connection into the ReLu function. The findings of the experiments indicate that this model has an overall recognition accuracy of 86.1%, which substantiates the claim that it is successful. Following the completion of this model's training, we developed and deployed the WeChat applet of agricultural disease and insect pest identification. After that, we began the real testing process. The findings demonstrate that the system is capable of correctly identifying crop illnesses and providing the appropriate recommendations.