基于数码相机的水稻氮素营养诊断

Nitrogen Status Diagnosis of Rice Based on a Digital Camera

作物叶片的颜色与其氮素营养状况密切相关,因此可以用叶片的绿色程度来反映其氮素水平。为了量化水稻叶片颜色特征并建立其与氮素营养状况间的关系,使用数码相机拍摄了不同品种、氮肥用量的水稻叶片图像,并比较了3个图像特征参数色相(H)、明度(V)、深绿色指数(DGCI)与SPAD值及水稻叶片含氮量的关系。结果表明,H、V、DGCI与SPAD值间存在良好的线性关系,拔节期、孕穗期DGCI和SPAD值间的决定系数分别为0.62**、0.60**。同时,3个特征参数和叶片含氮量间也存在良好的线性关系,孕穗期H、V、DGCI与叶片含氮量间的决定系数分别为0.53**、0.63**、0.59**。利用图像特征参数对水稻进行氮素营养诊断时选择孕穗期较好,3个特征参数中V较稳定,与水稻氮素营养间关系最好。

Greenness of leaves can reflect nitrogen level of crops due to the close relationship between teat color and nitrogen （N） content. The objective of this study was to quantify rice leaf greenness and reveal its relationship with N status through digital camera and image processing technology. We utilized a digital camera to take rice leaf photos of different varieties and N rates, and then compared the image feature parameters like hue （H）, value （V） and dark green color index （DGC1） with SPAD readings and leaf N content. Results showed good linear correlation between the H, V, and DGCI parameters and SPAD readings. The r2 between DGCI and SPAD readings was 0.62＂＂ and 0.60＂＂ at shooting and booting stage, respectively. The image feature parameters were also linearly correlated with leaf N content. The r2 between leaf N content and H, V, DGCI were 0.53＊＊, 0.63＊＊, 0.59＊＊ at booting stage, respectively. Booting stage was more proper than shooting stage while assessing the rice N status, and V was most stable among the three parameters, thus was optimal indicator.