Design Of Crop Leaf Disease And Insect Pest Identification And Monitoring System

China is a big agricultural country,daily life is inseparable from agriculture,but crop diseases and insect pests have a huge impact on agriculture,in the guarantee of environmental pollution reduction and high food quality at the same time,the prevention and control of crop diseases and insect pests is a very important work of agricultural staff.But crop species on the earth,crop species is very complex,in the field of the staff is not efficient and accurate to crop disease type judgment,if can’t accurately determine crop disease types,cause plant diseases and insect pests,area enlarges to prevention and control will delay affect crop yields,so determine crop species,Timely prevention and treatment is very important.In order to solve the above problems,this paper combined image recognition technology,deep learning technology of image feature extraction,embedded technology and wireless communication technology,etc.,and designed the recognition and monitoring system of crop leaf diseases and insect pests based on Android development platform.The main research contents are as follows:Firstly,this paper combines embedded technology with OV5640 camera controlled by STM32 for real-time monitoring of crop growth and surrounding environment,and transmits image data to Android client through gateway and central server through LoRa wireless communication technology.Then,combined with image recognition technology based on deep learning and PyTorch framework platform,the residual network Res NET-152 was used as the pre-training model,and Adam algorithm was used as the model optimization method to build the identification model of crop leaf diseases and insect pests.The Android client sends image data to the central server and uses the trained model to identify the picture of crop leaves to determine whether there is disease and the type of crop disease,providing support for the prevention and control of crop diseases and insect pests.Based on the Android platform was designed and implemented of crops diseases and insect pests image intelligent recognition application,application function is divided into two parts,the first is the real-time identification of crops diseases and insect pests species,the second is by monitoring the video status of crops and the surrounding environment,if the plant diseases and insect pests of crops,can to prevention and control of crop,Let our country agriculture more science and technology,information.