利用图像数字化处理技术，建立元帅系苹果叶片氮磷钾营养元素估算模型。选择不同营养水平的元帅系苹果叶样，通过计算机提取采集叶样的图像特征红R、绿G、蓝B各参数值及其相对的系数r、g、b和简单算术组合。分析RGB、rgb和简单算术组合分量值与叶片氮、磷、钾之间的相关特征。结果表明，叶片颜色特征分量G/B与氮元素含量呈极显著相关性，g分量与磷元素含量呈显著相关性，G/R分量与钾元素含量呈极显著相关性。选择叶片颜色特征分量跟元帅系苹果叶片氮、磷、钾元素的最高显著相关特征参数，建立的元帅系苹果叶片氮、磷、钾营养水平估算模型为y=-6.780x +8.913、y=-1.813x + 1.158、y=-25.196x + 22.872，其中x分别为叶片颜色特征G/B、g、G/R的分量值。氮、磷、钾预测值与实测值相关系数分别达到极显著(0.721)、显著(0.522)、极显著(0.644)水平，能较为可靠的反映元帅系苹果叶片的营养水平。

The estimation model of nitrogen， phosphorus and potassium nutrients in Red delicious apple leaves was established by image digital processing technology. The leaf samples of Red delicious apple with different nutritional levels were collected， and the parameter values of red R， green G and blue B， their corresponding coefficients r， g， b and simple arithmetic combination were analyzed by computer. The correlation characteristics between RGB， rgb and simple arithmetic combination component values and leaf nitrogen， phosphorus and potassium were calculated. The results showed an extremely significant correlation between leaf color characteristic component value G/B and leaf nitrogen content， G/R component value and leaf potassium content， and g component value was significantly correlated with phosphorus content. The most significant correlation characteristic parameters between leaf color characteristic components and leaf nitrogen， phosphorus and potassium elements in Red delicious apple leaves were selected. The estimation models of nitrogen， phosphorus and potassium nutrient levels in Red delicious apple leaves were established as y=-6.780x + 8.913， y=-1.813x + 1.158， y=-25.196x + 22.872， in which x was the component values of leaf color characteristics G/B， g， G/R， respectively. The correlation coefficients between the predicted value and the measured value of nitrogen， phosphorus and potassium which reached 0.721 (extremely significant)， 0.522 (significant) and 0.644 (extremely significant)， respectively， can reflect the nutritional level of Red delicious apple leaves reliably.

S661.1

天水市科技支撑项目“天水花牛苹果叶片营养分析应用研究”(2020-NCK-7871)。