Monitoring And Application Of Plant Nitrogen Nutrition Based On Multispectral Specificity Sensor

With the development of modern precision agriculture in China,real-time mastering of crop growth has become the key to realize precision agricultural management,and non-destructive and rapid monitoring of crop growth information using spectral technology is a prerequisite for realizing precision management in modern agriculture.Therefore,the development and exploitation of real-time spectral monitoring sensors for crop phenotypes is crucial for the sustainability,stability and long-term development of China’s agriculture.Currently,there are a wide range of spectral sensors on the market,however,the vast majority of them are mainly foreign products,which are expensive and not available for secondary development.In view of this,this study designed and developed a CNDM02 multispectral nitrogen nutrient diagnostic instrument,which has a good correlation with leaf nitrogen concentration and can provide direct feedback on crop leaf nitrogen content.At the same time,the device is lightweight and networked for transmission,which can be widely used in agricultural settings such as fields or greenhouses.The main findings of this thesis are as follows:（1）In view of the advantages and disadvantages of current multispectral nitrogen diagnostic devices,the design and development of a multispectral sensor-based nitrogen nutrient diagnostic device（CNDM02）was carried out based on years of research by the team.The device mainly consists of four parts:controller module（ESP8266）,data acquisition module（1 optical component）,power circuit module and housing.Arduino is used for programming and design,combined with China Mobile IOT platform,to realize the functions of remote collection,transmission and display of crop spectrum information.The debugging results show that the device can accurately acquire the potato leaf spectral information and has a good correlation with the leaf nitrogen concentration.Compared with similar products,CNDM02 can be integrated into precision agriculture system to realize real-time monitoring of crops and solve many problems such as difficult crop data collection.（2）In this study,the CNDM02 multispectral nitrogen diagnostic instrument was used to collect spectral data from potato leaves by conducting experiments at key fertility periods of potatoes.The purpose of this thesis is to design a spectral sensor to collect spectral data of potato leaves and correlate it with leaf nitrogen content.By using two-band,three-band and fused vegetation spectral indices for the analysis,we found that the CNDM02 multispectral nitrogen diagnostic instrument has good correlation with leaf nitrogen content.We investigated this correlation by constructing and validating two-band,three-band,and fused spectral index models.The experimental results showed that the overall performance of Opt-NDVI,Opt-RVI,Opt-CI_red-edge,Opt-SAVI and Opt-OSAVI was good in the two-band indices（R² in 0.63-0.64,RMSE in 0.54%-0.55%and RE in 10.10%-10.38%）.In the three-band and fused vegetation spectral indices,most of the indices performed well overall,and their estimation accuracy was not significantly different from that of the two bands,falling within the range of the two bands.The constructed model has a good estimation effect for potato leaf nitrogen monitoring.These results show that we successfully designed and developed a multispectral sensor-based crop nitrogen diagnostic monitor,which is fast,nondestructive,portable,highly integrated,networkable and controllable and low-cost.The CNDM02 spectral sensor combined with the spectral index model can be successfully used for potato leaf nitrogen content diagnosis.It provides an effective technical tool for agricultural production and can improve the yield and quality of potatoes.