摘要:本文利用手机相机获取玉米6叶期和9叶期的冠层图像，对图像进行色彩参数的提取与处理，分析了不同生长时期、不同品种间色彩参数的差异性，以及色彩参数与传统玉米氮素营养指标的相关性，选择出适宜的敏感色彩参数，对色彩参数与氮素营养指标进行拟合建模，建立了玉米氮素营养诊断体系，并推荐了不同产量目标下的施肥量，为实现利用智能手机田间拍照进行氮素营养诊断和精准推荐施肥提供参考。结果表明，在玉米6叶期，冠层图像色彩参数与传统氮素营养指标间的相关性优于9叶期，可作为应用数字图像分析技术进行氮素营养诊断的诊断时期；不同品种玉米的冠层图像色彩参数间无显著差异。B/（R+G+B）和G/（R+G+B）与传统氮素诊断指标——叶片SPAD值、第1完全展开叶叶脉硝酸盐浓度均显著相关，且B/（R+G+B）更为敏感，因此可作为玉米氮素营养诊断的色彩参数指标，诊断方程为：玉米叶脉硝酸盐浓度=1.73×1010×[B/（R+G+B）]9.43。并依此给出了不同B/（R+G+B）值下的玉米营养状况以及不同目标产量下的推荐施氮量。本研究结果可为基于手机相机开展玉米氮素营养诊断与推荐施肥技术的推广与应用提供技术支撑。
关键词:手机相机/
叶片数字图像/
色彩参数/
氮素营养诊断/
精准施肥/
夏玉米
Abstract:To meet requirements of food supplies and the accompanying pollution problem on environment, precision fertilization is one of the most important technologies. Soil nutrient test and crop nutrition diagnosis are essential work for precision fertilization. With the current situation of the increasing agriculture scale management, it is urgent to develop fast, nondestructive and economic techniques for the nitrogen nutrition diagnosis of crops. Digital images technology has been widely applied in nutrition diagnosis of crops. In majority of such researches, digital cameras have already been successfully used. However, few researches were reported to use cellphone cameras to study nutrition diagnosis and precision fertilization of crops. Thinking of the advantages that cellphone cameras have, such as portability, universality and handleability, the application of cellphone cameras should be detailly studied in nutrition diagnosis. In this study, we used smart cellphones to photograph corn leaves at 6-leaf and 9-leaf stages. The color parameters of corn leaves images were extracted and processed. The differences in color parameters of leaves photographs during two growth stages and for four varieties of corn were evaluated. The correlations of parameters with traditional nitrogen nutrient indexes were determined. Appropriate color parameters were selected based on statistical analysis and nutrient diagnosis model established for the color parameters and nitrogen nutrition index. Then the model was fitted to establish indicator systems of diagnosis of nitrogen nutrient and recommendations for fertilization of corns. The results showed that correlations of color parameters and nitrogen nutrient indexes at 6-leaf stage were more significant than those at 9-leaf stage, suggesting that 6-leaf stage was suitable time for diagnosis of corn nitrogen nutrient using digital image processing technique. From the analysis of leaves photographs of four corn varieties, there was no statistically significant difference among the images. Furthermore, the consequences supported two color parameters, B/(R+G+B) and G/(R+G+B) as candidates for sensitive color parameters. These two color parameters both had strong correlations with leaf SPAD and vein nitrate concentration. Also based on multivariate analysis, B/(R+G+B) was the best and was selected as sensitive color parameter for diagnosis of corn nitrogen nutrient. The diagnosis model of vein nitrate concentration was 1.73×1010×[B/(R+G+B)]9.43. Based on the equation, nitrogen application rates under different B/(R+G+B) values were calculated for certain yield targets of corn. The results were applied to nitrogen nutrient diagnosis and recommendation of fertilization of corn. In summary, it was possible and applicable to take photographs of corn leaves at 6-leaf stage with smart cellphone, extract B/(R+G+B) color parameter and use it to diagnose nitrogen nutrition status.
Key words:Cellphone camera/
Leaf digital photography/
Color parameter/
Diagnosis of nitrogen nutrient/
Precision fertilization/
Summer corn
PDF全文下载地址:
http://www.ecoagri.ac.cn/article/exportPdf?id=a47c365d-8625-4403-9df5-c3ecfd990b3d
