中文摘要：

为探索基于数字图像处理技术的冬小麦氮素无损诊断图像评价指标及构建方法,设计拍摄2012-2014年度不同种植方案下冬小麦冠层图像,基于归一化的H分量K均值聚类分割算法提取基础颜色特征值,与同期叶片氮含量（leaf nitrogen content,LNC）进行线性拟合,调优并确定三原色分量最佳拟合系数,提出RGB空间下的颜色组合标准化指数（normalized color mix index,NCMI）。对比深绿色指数（dark green color index,DGCI）、红光标准化值（normalized redness intensity,NRI）和绿光与红光比值G/R发现,3个采样期NCMI与LNC的决定系数R~2均高于3个对比指标,分别为0.77、0.79、0.94,均方根误差（root mean square error,RMSE）相较同期最低的指标,分别降低了0.18%、0.37%和1.67%;生选6号和扬麦18号NCMI与LNC的相关性,在一定冠层覆盖度下均优于其他3个指标;D2密度（3×10^6株/hm~2）N1（纯氮150 kg/hm~2）处理下NCMI效果明显优于其他3个指标,R~2和RMSE较NRI分别改善了7.69%和4.11%,该研究可为一定冠层覆盖度下的冬小麦氮素营养诊断图像评价指标提供参考。

英文摘要：

This paper explored the digital image evaluation index of the monitoring of winter wheat nitrogen nutrition and the establishment method based on digital image processing technology. The images of winter wheat canopy at different stages under different planting schemes （2 varieties, 2 planting density levels and 3 nitrogen fertilizer application rates） were taken from 2012 to 2014, and the basic color characteristic value of canopy image samples was extracted using the algorithm based on the normalized k-means clustering segmentation for theH component. Firstly, it chose 9 image feature parameters, including 3 monochromatic components （r, g andb）, 3 linear combination parameters （r-g-b,r-g andr-b）, and 3 linear combination parameters by standardized processing （（r-g-b）/（r＋g＋b）, （r-g）/（r＋g＋b） and （r-b）/（r＋g＋b））, and analyzed contrastively the correlation between the above parameters and the monitoring evaluation index of nitrogen nutrition under different schemes. It was found that the characterization ability of 3 monochromatic components for wheat nitrogen nutrition level was not the same, but for the combination characteristic parameters by the 3 monochromatic components, the degree of correlation was better than the 3 monochromatic components, and at the same time, the increase of the level of correlation was more obvious after further normalized. So, to establish the image evaluation index of the monitoring of wheat nitrogen nutrition, the first step was to choose the base color component which had a stronger representation ability of leaf nitrogen content and increase its proportion, the second step was to adjust and optimize the combination weights of the other remaining monochromatic components, and the final step was to standardize it. A linear fitting was carried out at different sampling periods and under different cultivation schemes, which made use of basal image color characteristic values and leaf nitrogen concentration （LNC） measurement during