A Research On Dynamic Meteorology Model And Remote Diagnosis Method For Wheat Diseases And Insect Pests Based Oninternet Of Things

Wheat is one of the main grain crops in China. The plant diseases and insect pests are the important biological disasters affecting wheat yield. Grasping the occurrence and development of wheat diseases and insect pests in time is of great significance for the interests of the agriculture operators and the national food security. Under the theme of dynamic monitoring of wheat diseases and insect pests, based on the existing Central Monitoring Platform of IoT of Wheat, the dynamic meteorology model and the remote diagnosis method for wheat diseases and insect pests were designed.According to the occurrence and epidemic characteristics of wheat in the Huanghuai region and the response characteristics of climatic conditions, we set up the early warning model of prevalence probability for stripe rust with two main meteorological data of the air temperature and air humidity. By integrating the early warning model into the Central Monitoring Platform of IoT of Wheat and using the real-time acquisition of meteorological data from the IoT field monitoring stations, combining the wheat variety data and the result of remote diagnosis for wheat diseases and insect pests, the early warning model could response well to the basic conditions of the field, and do the good work on early warming for stripe rust of wheat.Remote diagnosis method for wheat diseases and insect pests was based on image recognition technology basic method. Based on the Central Monitoring Platform of IoT of Wheat, the application system which integrating the image acquisition, image recognition and image diagnosis was developed. By the image remote-acquired from the Platform, the classification and identification of four kinds of wheat leaf pictures which had got powdery mildew, rust disease, aphis infected and healthy ones were realized. After then a comparative experiment on each kind of wheat leaf pictures by means of image segmentation, feature extraction and digital image classification was conducted. The test results showed that the recognition rates reached desired levels. Among these results, the recognition rate of powdery mildew was 82.5%, the recognition rates of rust disease, aphis and healthy leaves were all above 95%.This study also used multi-source data coupling technology, which increased the effect of monitoring for wheat diseases and insect pests. Dynamic meteorology model and remote diagnosis method for wheat diseases and insect pests based on IoT had good practicability and extensibility, which made an innovative important step for the application and development of Internet of Things in the field of agriculture.