基于手机相机获取冬小麦冠层数字图像的氮素诊断与推荐施肥研究

本文利用不同型号手机、通过不同拍摄角度获取冬小麦拔节期冠层图像,并对其图像进行色彩参数的提取、处理与分析,与传统小麦氮素营养指标进行相关性分析,筛选出敏感色彩参数,对二者进行拟合建模,建立了冬小麦氮素营养诊断指标体系和推荐施肥指标体系,为作物精准施肥提供参考。研究结果表明,在获取冬小麦冠层图像时,适宜从逆光俯视的角度拍摄,不同型号的手机拍照获取的冠层图像色彩参数没有明显差异,冠层图像色彩参数中可见光大气阻抗植被指数(VARI)及红光标准化值R/(R+G+B)]与传统诊断指标叶片SPAD值、茎基部硝酸盐浓度均有显著的相关关系;其中VARI最为敏感,可作为冬小麦氮素营养诊断的色彩参数指标,诊断方程为冬小麦茎基部硝酸盐浓度=1.481×106×VARI4.987,依据此给出了不同VARI值下的冬小麦营养状况以及推荐施氮量。并基于此研究成果进行了手机软件开发,建立了一款针对冬小麦氮素营养诊断与推荐施肥的软件,为基于手机相机开展冬小麦氮素营养诊断与推荐施肥技术的推广与应用提供了技术支撑。

Study on nitrogen nutrition diagnosis and fertilization recommendation of winter wheat using canopy digital images from cellphone camera

Digital cameras have been successfully used in nutrition diagnosis of crops. Few researches were reported on the application of cellphone cameras on nutrition diagnosis and precision fertilization of crops, though the cellphone cameras have various advantages, such as portability, handleability and universality. In this study, we used Samsung and Xiaomi smart cellphones to photograph winter wheat canopy at jointing stage from four shooting angles, frontlighting looking down, frontlighting overlooking, backlighting looking down and backlighting overlooking. The color parameters of achieved winter wheat canopy images were extracted, processed and analyzed. And their correlations with traditional nitrogen nutrition indexes were analyzed. According to the statistical analysis of the correlation, appropriate color parameters were selected, and the nutrition diagnosis model was established with color parameters and nitrogen nutrition index. Then the model was fitted to establish indicator systems of diagnosis of nitrogen nutrition and recommendation system of fertilization for winter wheat. The aim of the study was to provide references for application of smart cellphone on precision fertilization of crops. The results showed that there was no remarkable difference in the color parameters of canopy images taken with different types of smart cellphones. However, a certain difference was observed in color parameters of canopy images taken from different shooting angles. More than half of color parameters of canopy images taking from the view of backlighting overlooking were significantly correlated with traditional nutrient indexes. Furthermore, the results indicated that the color parameters VARI(G-R)/(G+R-B)] and R/(R+G+B) both had outstanding correlations with leaf SPAD and stem nitrate concentration. And VARI was found to be the best, and therefore, selected as the sensitive color parameter for winter wheat nitrogen nutrition diagnosis. The diagnosis model was stem nitrate concentration=1.481×10⁶×VARI^4.987. According to the equation, the nitrogen nutrition status of winter wheat was normal when VARI was between 0.201 3 and 0.250 9. The nitrogen application rates under different VARI values were calculated for different target yields of winter wheat. The results were applied to develop cellphone software for nitrogen nutrition diagnosis and fertilization recommendation of winter wheat. In summary, it was possible and applicable to take photographs of canopy from the view of backlighting overlooking with smart cellphone and extract VARI color parameter to diagnosis nitrogen nutrition status. The results provided technical support for the diagnosis of nitrogen nutrition and recommendation of fertilization of winter wheat based on cellphone camera.