| Precision crop management in the process of crop production is the important part of precision agriculture which is is the inevitable choice for sustainable agriculture.The core idea is farming precise control of water and fertilizer according to positioning precision the difference of crop growth status,in order to reduce the costs,improve production efficiency,and reduce environmental pollution.While,the crop growth information real-time acquisition is based of precision crop management,play a decisive role not only to ensure the quality,yield of crop but also the increase of farmers' income and production.Nowaday,crop growth information monitoring technology based on the leaf canopy reflectance spectra is the main technology of precise management of crop growth,and the crop growth sensor as a core part of perceptual information,decides the quality of crop canopy spectral information acquisition.Affects the generation of accurate fertilization scheme directly.In recent years,Some research on crop growth monitoring sensor based on crop growth monitoring from the domestic and foreign,but there is no practical and commercial device,which could collect the crop growth information in real-time,on-line,which seriously restrict the development of precise management technology of crop production,so there is an urgent need for the development of portable crop equipment with a high rate of performance and price,wide measuring range growth monitoring and diagnosis,in order to solve the bottleneck of long cycle in field crop growth information detection,couldn't guide farming timely and effective.The research has just started in our country,so far no device corresponding commercial,even in the absence of device which integrated crop growth information acquisition timely and fastly.There is an urgent need to study a portable monitoring device based on embedded technology which included intelligent crop detection method,to solve the bottleneck of crop growth information monitoring,providing technical support equipment for precision crop management in informationalization and intellectualization.In this paper,Based Crop Growth Monitoring and Diagnosis CGMD302 developed by National Engineering Technology Center for Information Agriculture of Nanjing Agricultural University as research object.Optimize of the structure and function according to the actual situation of field application.Mainly includes:optimize the calibration method of crop growth sensor;optimize the sensor signal conditioning circuit for crop growth,optimize the application functions of CGMD302,as follows:1.Analysis the problem of crop sensor by original calibration method,propose a radiometric calibration method based the standard detector to calibrate crop sensor,built a automatic calibration platform by integrating sphere with uniform light source and standard luminance meter(Traced to National Institute of Standards and Technology,NIST),used to calibrate the crop growth sensor's DN to the brightness/radiation illuminance,the time for one sensor complete calibration reduce from 40minutes to 10 minutes,it has nothing to do with weather.The error from calibration reduced from +50.13%to 0.83%,average error caused by the environment light decreased from 8.3%to 1.22%.BY the optimization of instrument calibration method,not only improve the calibration speed and accuracy,but also enhance the environmental adaptability of the luminance.2.Analysis the defects of the T type amplifying circuit in the weak light current signal processing,design of the integral circuit with gain adjustable,can achieve a wide range of spectral information and high precision nondestructive sensing.The spectral reflectance of the circuit after processing,the resolution in the light radiation intensity of Ep(815nm)>30mW/(nm·m~2),better than 0.1%(the T type amplifying circuit in same condition is 9.1%).3.Increase the diagnostic function.Through multiplexing,design a concise man-machine interface which convenient parameter input and LCD display with limited peripheral.Mixed the parameters and the diagnosis model together,realize the recommended fertilization information output directly.4.In the field experiment(wheat and rice),Compare a multi spectral sensor with Fieldspec Pro FR2500 spectrometer.The reflectance measured by CGMD302 are similar to reflectance monitoring by ASD.The optimized CGMD302 has the average error of 730nm channel and 815nm channel is 2.18%and 3.64%respectively,better than original CGMD302,the root mean square error is 2.62%,4.47%respectively,reduce from original CGMD302 from 5.48%and 8.61%;what's more,the ratio of vegetation index(RVI)and normalized difference vegetation index(NDVI)of rice and wheat measured by CGMD302 and ASD has a strong correlation,the correlation coefficient was 0.93,0.91 respectively.The relative root mean square error(RRMSE)between CGMD302 and ASD on vegetation index(RVI,NDVI)were 0.0649 and 0.0884 respectively;The correlation between Crop Agronomic parameters and vegetation index obtained by CGMD302 were good,such as leaf nitrogen content(LNC),leaf nitrogen accumulation(LNA),leaf area index(LAI),leaf dry weight(LDW)correlation coefficient were 0.83,0.745,0.735,0.739 respectively.The average error of LNA,LNA,LAI,and LDW is 11.11%,9.13%,10.37%,5.58%respectively.The experimental results show that the optimized CGMD302 had a better performance in precision,stability and resolution in reflectivity than the original one,which would better meet the precision and the rapid of crop growth information acquisition request. |
PDF Full Text Request