Quantitative Remote Sensing Monitoring And Control Of Rice Blast

Ensuring food security is the primary task of Rural Revitalization Strategy.China’s food industry is facing a new situation and pressure from home and abroad.It has turned into an urgent demand to build a strategic policy system of food security,ensure food supply and security,and firmly hold "China’s rice bowl" in their own hands.In grain production,agricultural pests and diseases are the primary factor resulting in yield loss.The large-scale outbreak of pests and diseases not only affects the enthusiasm of farmers to grow grain,but also restricts the high-quality development of the grain industry.Rice is the most important grain crop in China,and the occurrence of rice blast is high,which seriously threatens the sustainability of rice production.As one of the Class I diseases and insect pests,rice blast has the characteristics of fast transmission speed and great difficulty in prevention and control,especially panicle neck blast has the greatest impact on rice yield.At present,variable pesticide application is the most effective method for rice blast prevention and control.The use of unmanned aerial vehicle hyperspectral remote sensing platform can realize the identification,monitoring,prediction and early warning of the incidence of rice panicle blast and cervical blast at the field level,and provide prescription map for variable topdressing and application of rice blast by unmanned aerial vehicle.Based on the field experiment of rice blast in Wanchang Town,Yongji County,Jilin City,this study used the UAV hyperspectral remote sensing platform to obtain the images of the whole growth period of rice,and collected the data of disease grade and C/N ratio of rice in 30 different areas on the ground.Four different parameters were used as inputs to establish the random forest classification model,in which the combination of vegetation indices achieved the best result.Based on the random forest model,the disease situation of rice in the field environment was monitored and inversed,and the early warning prediction of rice blast was preliminaries explored in combination with meteorology and rice carbon C/N ratio.Finally,according to the early warning and monitoring results,the fertilization,pesticide application and management countermeasures were put forward.The results show that:(1)There were characteristic differences in visible and near-infrared spectral curves of four rice varieties with different degrees of rice panicle blast.The original spectral curves of rice samples with different degrees of disease were mainly affected by chlorophyll changes,and showed significant differences at 550 and 670 nm.In the red edge band of 670-760 nm,the location and amplitude of the curve changed under the stress of rice panicle blast.The reflectance of near infrared band decreases with the aggravation of the disease.(2)The results show that the random forest model has superior accuracy and strong consistency in the classification of rice panicle blast.The precision of the random forest model with the combination of vegetation indices as input is 90%,and the kappa coefficient is 0.86,which indicates that the model has a strong "balance".(3)The hypersensitive platform of UAV can realize the monitoring and early warning of rice panicle blast at field level,and promote the revolution of agricultural technology.UAV hyperspectral remote sensing realized the mapping of the early incidence of rice panicle blast in different areas of the whole field,and the accuracy reached 82.25%,kappa coefficient was 0.743;combined with the meteorological information of the IOT,the early warning of rice blast could be realized at the same time.(4)Farmers’ unreasonable field operation management is the main external cause of rice blast.Blind overuse of pesticides increases the planting cost.Information blocking is one of the obstacles to predict and control the outbreak of rice blast.This study is aimed at the accurate monitoring,prediction and early warning of rice blast,which can improve the modernization level of rice planting,promote stable and increased grain yield,increase farmers’ enthusiasm and income,and help the implementation of smart agriculture,food security strategy and Rural Revitalization Strategy.