Monitoring Research On Rice Growth In Northeast Cold Region Based On UAV-Borne Spectral Imager

As the most populous country in the world,China ’ s basic national policy is to protect people’s livelihood and national food security.Rice is the main grain economic crop in China,and the safety of grain production is related to the stability and sustainable development of Chinese society.However,as the most serious disease of rice,rice blast has the characteristics of strong infectivity,wide occurrence area and difficult to control.With the vigorous development of ’ smart agriculture and precision agriculture ’,the use of UAV remote sensing technology to monitor crop growth has been paid more and more attention.UAV remote sensing solves the problem of inefficient and time-consuming information acquisition of large-scale crops.At the same time,the development of crop remote sensing technology further needs to use spectral technology to solve the problem of timely and real-time dynamic monitoring of pests and diseases.In this paper,the UAV-borne multispectral imaging detection technology was used to study the growth of rice and the spectral image characteristics of rice blast in the northeast cold region.Based on the structure of rice canopy and single leaf,the spectral scattering mechanism of rice canopy and single leaf was studied.The spectral characteristics of rice canopy reflection were analyzed and the band selection was carried out.The statistical analysis method of health and disease degree by spectral index was proposed.The data acquisition and preprocessing of UAV-borne multispectral imaging are studied.According to the parameters and functions of UAV-borne multispectral equipment,the flight and acquisition methods of test field and unmanned airborne equipment are planned.The obtained spectral images are preprocessed by optical lens nonuniformity correction,atmospheric correction,geometric correction and other spectral images.The multi-spectral detection experiment of rice canopy was carried out,the data selection was carried out,and the correlation between ground and UAV-borne spectrum was analyzed.Based on the UAV-borne spectrum data of three groups of regions at 50 m altitude and four growth periods of eight plots in each group,six vegetation indexes were used to invert the rice blast grade.The results showed that in the tillering stage and jointing stage of rice growth,the leaves were in the growth stage,and the remote sensing images of RVI and NDVI could be used to judge the trend of rice blast.In the booting stage of rice growth,the canopy coverage of rice leaves is high,and the discrimination between RVI and NDVI is reduced.The rice blast damage will destroy the internal tissue structure of rice leaves,and the accuracy of the constructed BNDVI is improved,which can be used as the basis for the mid-term judgment of rice blast disease.In the filling stage of rice growth,rice leaves were damaged in large areas due to rice blast,and NDRE was used as the basis for the judgment of rice diseases in the later stage.Finally,the values of remote sensing images at different rice growth stages after vegetation index transformation are input into the linear discriminant analysis(LDA)and decision tree classification model as parameters,which can better classify the degree of rice blast.