中文摘要：

传统的水稻氮营养诊断需要大量的破坏性取样和实验室分析，SPAD叶绿素仪虽然具有较高的精度，但只能测定叶片的很小一部分，需要进行多点的测试，在大面积应用费时费力，时效性和便捷性不足。该文利用田间不同氮水平试验，使用数码相机获取了水稻冠层图像并进行色彩分析，研究了利用可见光遥感技术手段进行水稻冠层的氮营养诊断的可行性，并对表征水稻氮营养状况的可见光光谱诊断指标进行了探索。研究结果表明，数字图像红光值（R）、绿光值（G）和红光标准化值（NRI）与水稻氮营养状况常规诊断指标植株全氮含量、生物量和地上部吸氮量等都呈显著的线性反相关关系，而绿光标准化值则与上述指标望显著正相关关系。红光标准化值与叶绿素仪SPAD读数相比能更好地表征水稻幼穗期的氮素营养状况，与常规测试指标的相关系数介于0．541～0．780之间。数字图像色彩分析技术具有发展成为新一代作物氮营养诊断手段的潜力。

英文摘要：

In the present research, a field experiment with different N application rate was conducted to study the possibility of using visible band color analysis methods to monitor the N status of rice canopy. The Correlations of visible spectrum band color intensity between rice canopy image acquired from a digital camera and conventional nitrogen status diagnosis parameters of leaf SPAD chlorophyll meter readings, total N content, upland biomass and N uptake were studied. The results showed that the red color intensity （R）, green color intensity （G） and normalized redness intensity （NRI） have significant inverse linear correlations with the conventional N diagnosis parameters of SPAD readings, total N content, upland biomass and total N uptake. The correlation coefficient values （r） were from --0. 561 to --0. 714 for red band （R）, from --0. 452 to --0. 505 for green band （G）, and from --0. 541 to 0. 817 for normalized redness intensity （NRI）. But the normalized greenness intensity （NGI） showed a signifi- cant positive correlation with conventional N parameters and the correlation coefficient values （r） were from 0. 505 to 0. 559. Compared with SPAD readings, the normalized redness intensity （NRI）, with a high r value of 0. 541-0. 780 with conventional N parameters, could better express the N status of rice. The digital image color analysis method showed the potential of being used in rice N status diagnosis in the future.