Research On Huang Huazhan Rice Fertilization Decision Model Based On Remote Sensing Of Unmanned Aerial Vehicles

China is a big agricultural country.Rice is one of the most important food crops in China.Stable and high-yielding rice is essential for ensuring food security in China.Nitrogen is one of the most effective fertilizers for increasing crop yields.The traditional fertilization method mainly relies on empirical fertilization,resulting in uncoordinated proportion of nutrients in fertilization,high input,low yield,and the quality of crops declines,and a large amount of nitrogen fertilizer is not effectively absorbed and utilized,and will be lost in some form.It also has an impact on the ecological environment that cannot be ignored.The precise management of nutrients is one of the important means to ensure the healthy,high-quality and high-yield rice and to protect the agricultural ecological environment.This study is aimed at the current situation of excessive application of nitrogen fertilizer in rice in China.Taking Huanghuazhan as the research object,three different regions of Guangzhou(test area 1),Lechang(test area 2)and Hunan Yiyang(test area 3)were selected.Nitrogen nutrition diagnosis technology and fertilization status of rice growth period at home and abroad,using SPAD chlorophyll meter,Green Seeker RT100 and rededge-M multispectral camera to obtain field data of rice growth period in each experimental area,and analyzing three different sensors for diagnosis of rice nitrogen nutrition The level and its difference,the fertilization model of rice key fertilization node based on UAV remote sensing technology is established,which provides a fast,accurate and large-area technical method for rice nitrogen nutrition monitoring and variable fertilization.It is of great significance for rational and effective fertilization,increasing the utilization rate of nitrogen fertilizer and reducing the pollution caused by fertilizer to the environment.The main work and conclusions of the study are as follows:(1)The variation of the data measured by each sensor in the whole growth period of rice in each region and its correlation with the nitrogen content of rice plants were analyzed.The results showed that the SPAD value showed a trend of “first rising,then decreasing,then rising and then decreasing” during the whole growth period of rice.The correlation coefficient between SPAD value and nitrogen content of rice plants in all regions was above 0.72,which was significantly correlated with Lechang.The coefficient is the highest,followed by Yiyang.Guangzhou’s lowest,NDVIg(Green Seeker sensor measurement)and NDVIa values(unmanned aircraft equipped with rededge-M multi-spectral camera)showed a trend of “first sharply rising,then flattening,then slowly decreasing” during the whole growth period of rice.The correlation coefficient between NDVIg and rice plant nitrogen content in all regions was above 0.6,Yiyang was the highest,followed by Guangzhou and Lechang.The NDVIg value is highly correlated with the NDVIa value and the SPAD value.The fitting coefficient R2 has a mean value of 0.9 or more.Therefore,this paper optimizes the nitrogen nutrition status of rice by remote sensing technology.(2)The main flight operation parameters of the UAV remote sensing monitoring system are optimized,and the optimal flight altitude and working time period are determined.The results show that the NDVI values of different regions are consistent at different heights,and the NDVI values between 20-60 m are quite different.The NDVI values of 80 m and 100 m are not changed much,and the corresponding NDVI coefficient of variation is 0.027,indicating 80 m and The stability of the NDVI index measured at 100 m is better.The NDVI value measured at different time intervals from 6:00 am to 17:00 pm is "increasingdecreasing-increasing-decreasing" with time,and the correlation coefficient between NDVIa value and NDVIg value from 10:00 Stable at around 0.895.Therefore,the best flight height for the nitrogen nutrition status of rice using the rededge-M multispectral camera is 80m-100 m,and the best aerial time period should be 10:00-14:00.(3)The variable fertilization model of rice in different regions based on UAV remote sensing technology was established,and the differences of inter-regional models were analyzed.The results showed that the R2 of the NDVI index model of the rice in the different regions was: 0.887,0.981 and 0.985,respectively,indicating that the NDVI index of rice could be inverted by this model.The standard planting ratio index RISP composed of NDVIa and the standard nitrogen application rate was highly correlated with the nitrogen content of rice plants in each region.The determination coefficient R2 of each regional fertilization model based on the standard planting ratio method was 0.991 and 0.879,respectively.0.947.(4)The verification experiment of rice fertilization node variable fertilization model based on UAV remote sensing technology was carried out in Guangzhou.The results showed that the NDVI index of rice treated with nitrogen in remote sensing model was basically consistent with the NDVI index of standard nitrogen application.The yield is basically consistent with the standard nitrogen application;the fertilization decision model based on the remote sensing technology of UAV is used to accurately manage the nitrogen fertilizer of rice.While ensuring the yield,it can avoid the abuse of nitrogen fertilizer and reduce the cost input,making the application of rice nitrogen fertilizer more reasonable.And efficient.