Preliminary Study On Diagnosing Nitrogen Status Of Winter Rapeseed Based On Digital Image Processing Technique

Winter rapeseed is a dominant oil crop in China, and reasonable application of nitrogen fertilizer is important for the yield and quality. It is a significant part of nitrogen nutrition management that dressing timely and appropriately according to the nutritional status of crop growth process.With the development of the modern science and technology, digital image processing technique is the main direction of crop nitrogen diagnosis in recentital camera rapidly and non-destructively, and our study provides a theory and technique foundation of digital image processing technique on rapid and non-destructive nitrogen diagnostic of winter rapeseed. The main results were as follows: years. In this study, winter rapeseed was used in field experiments with different nitrogen applications rates in two years. The canopy pictures of winter rapeseed were obtained with digital camera, meanwhile analyze the correlations between the information of canopy pictures and conventional nitrogen diagnosis parameters, and then, determined the best period, diagnosis index and equation model by using digital camera to estimate the nitrogen concentration of winter rapeseed. On the basis of above, we standardized the operating range when acquiring the canopy pictures of winter rapeseed. The study determined best best period, diagnosis index, shooting parameter and equation model by using digital camera rapidly and non-destructively, and our study provides a theory and technique foundation of digital image processing technique on rapid and non-destructive nitrogen diagnostic of winter rapeseed. The main results were as follows:1. Digital image processing technique can be used in nitrogen nondestructive diagnosis of winter rapeseed, the stage before after transplanting 126 days are crucial period with NRI as the index for the nitrogen diagnosis. Compared to other digital parameters(R, G, B, G/R, G/B, B/R, NGI and NBI), it was more convenient between normalized redness intensity(NRI) and nitrogen application rate, aboveground biomass, leaf nitrogen concentration, chlorophyll in leaf, nitrogen uptake and nitrogen nutrition index. At Bud period, the normalized redness intensity(NRI) showed a prominently and very prominently inverse relationship with different equation model and the regression equation with nitrogen application rate, aboveground biomass, leaf nitrogen concentration, chlorophyll in leaf, nitrogen uptake and nitrogen nutrition index is y(t/ hm2)=-8.003x+2.706, y(t/ hm2)=-64.721x+23.071, y(g/ kg)=-6.930x+2.618, y(mg/ kg)=-12.750x+5.665, y(kg/hm2)=-4087.416x+1414.274 and y=-27.198x+9.812, the correlation values(R2) is 0.917**, 0.746** and 0.953** between measured and estimated leaf nitrogen concentration, chlorophyll in leaf and nitrogen nutrition idex, RMSE and RE were 0.821, 26.32%, 0.330, 28.57% and 0.228, 28.39%.2. The optimal operating range of digital image processing technique to estimate nitrogen nutrition index of winter rapeseed was followed: the best shooting time is sunny noon because of its relatively greater solar elevation angle and stable light, the image can be taken at any height near the ground, it is easy to operate at the image angle of 30°-60°, the shooting mode should be automatic exposure mode, the photo pixel is better in relatively clear and high resolution, and the images saving in compressed JPEG accurate format with less space occupying is recommended. NRI in different shooting parameters such as different time of a day, shooting angle, photo pixel, illumination, image format, shooting mode and shooting height can also get better results at different nitrogen applications. Our study provides a theory and technique foundation of digital image technique on non-destructive nitrogen diagnostic using low-altitude unmanned aerial vehicles remote sensing.3. Quadratic equation was fitted between nitrogen application rates and yield, and regression equation was y=-0.016x2+9.394x+1336.486. With the equation, we determined the highest theoretical yield was 2715 kg/hm2 with the corresponding nitrogen application rate was 294 kg/hm2. The critica NRI in Over-wintering period and Bud period were 0.3154 and 0.3177, and the equation model of topdressing were N(kg/hm2) =294-(NRI-0.3473)/(-1.312×10-4) and N(kg/hm2) =294-(NRI-0.3440)/(-1.067×10-4) respectively.