[1]刘文豪,徐俊增*,陆 韬,等.基于智能手机可见光成像的水稻叶片氮素状况诊断[J].江苏水利,2019,(04):37-41.
　LIU Wenhao,XU Junzeng*,LU Tao,et al.Diagnosis of nitrogen status of rice leaves based on visible light image of smart phone[J].JIANGSU WATER RESOURCES,2019,(04):37-41.
点击复制

基于智能手机可见光成像的水稻叶片氮素状况诊断()
分享到：

《江苏水利》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2019年04期
页码:
37-41
栏目:
水利信息化
出版日期:
2019-04-15

文章信息/Info

Title:
Diagnosis of nitrogen status of rice leaves based on visible light image of smart phone
文章编号:
1007-7839(2019)04-0037-05
作者:
刘文豪1;  徐俊增1*;  陆 韬2;  朱莉莉2;  王乙江3;  徐玉良4
1.河海大学农业工程学院, 江苏 南京 210098; 2.昆山市城市水系调度与信息管理处, 江苏 苏州 215300; 3.昆山市张浦水利(水务)站, 江苏 苏州 215300; 4.昆山市水利局, 江苏 苏州 215300
Author(s):
LIU Wenhao1;  XU Junzeng1*;  LU Tao2;  ZHU Lili2;  WANG Yijiang3;  XU Yuliang4
1.College of Agricultural Engineering, Hohai University, Nanjing 210098, Jiangsu; 2.Urban Water Scheduling and Information Management Department of Kunsha, Suzhou 215300, Jiangsu; 3.Kunshan Zhangpu Water Conservancy Station, Suzhou 215300, Jiangsu; 4.Kunshan Water Resources Bureau, Suzhou 215300, Jiangsu
关键词:
可见光图像;  符号回归算法;  颜色特征组合;  SPAD值;  智能手机
Keywords:
visible light image;  symbol regression algorithm;  color feature combinations;  SPAD value;  smart phone
分类号:
S511
文献标志码:
B
摘要:
稻田氮肥过量施用造成氮肥利用率低下,加剧了农田面源污染。实时实地的氮素亏缺诊断技术将为实地氮肥管理提供决策,可以合理控制施肥量,提高氮肥利用效率。经长期研究,叶绿素相对值(SPAD值)可以作为水稻氮素营养水平的准确反映。通过智能手机拍摄不同生育期各氮素营养水平下的水稻冠层可见光图像,利用Eureqa软件的符号回归算法进行不同颜色特征组合与SPAD值之间的关系拟合,分别建立了水稻返青期、分蘖期、拔节孕穗期基于可见光图像的最优SPAD值拟合模型。结果表明,验证期各模型相关系数(r)均在0.9以上; 均方根误差最低为4.82,最高为7.41; 平均绝对误差最低为1.87,最高为2.06,具备较高预测精度,总体上,拔节孕穗期模型精度最高。本研究旨在为基于智能手机的水稻氮素亏缺诊断与水稻实地氮肥管理提供决策支持。
Abstract:
Excessive application of nitrogen fertilizer in rice fields results in low nitrogen use efficiency, which exacerbates non-point source pollution in farmland.The real-time field nitrogen deficiency diagnosis technology will provide decision-making for site-specific nitrogen management, which can reasonably control the amount of fertilizer application and improve nitrogen use efficiency.After long-term studies, the relative value of chlorophyll(SPAD value)can be used as an accurate index of nitrogen nutrient condition in rice.The visible light image of rice canopy under different nitrogen application levels at different growth stages was photographed by smart phone.The symbol regression algorithm of Eureqa software was used to fit the relationship between different color feature combinations and SPAD values, and the models of rice turning green period, tillering period and jointing period were established respectively.The results showed that the correlation coefficient(r)of each model in the verification period was above 0.9; Among 3 models, the lowest root mean square error was 4.82, the highest was 7.41; the lowest average absolute error was 1.87, the highest was 2.06, and the prediction accuracy of all models was high.In general, the model of jointing and booting stage had the highest prediction accuracy.The research results could provide decision support for rice nitrogen deficiency diagnosis and site-specific nitrogen management based on smart phones.

参考文献/References:

[1] 张耗, 褚光, 剧成欣, 等.实地氮肥管理对水稻根系形态生理和产量的影响[J].扬州大学学报(农业与生命科学版), 2013, 34(04):62-66+93.
[2] 于艳梅.明沟控制排水稻田水氮流失规律与模拟[D].南京:河海大学, 2014.
[3] 王涛, 刘洋, 左月明.作物氮营养无损诊断研究进展[J].农业研究与应用, 2013(06):56-60.
[4] 吴良欢, 陶勤南.水稻叶绿素计诊断追氮法研究[J].浙江农业大学学报, 1999(02):27-30.
[5] 艾天成, 李方敏, 周治安, 等.作物叶片叶绿素含量与SPAD值相关性研究[J].湖北农学院学报, 2000(01):6-8.
[6] 王远, 王德建, 张刚, 等.基于数码相机的水稻冠层图像分割及氮素营养诊断[J].农业工程学报, 2012, 28(17):131-136.
[7] 李岚涛, 张萌, 任涛, 等.应用数字图像技术进行水稻氮素营养诊断[J].植物营养与肥料学报, 2015, 21(01):259-268.
[8] 贾良良, 范明生, 张福锁, 等.应用数码相机进行水稻氮营养诊断[J].光谱学与光谱分析, 2009, 29(08):2176-2179.
[9] 李刚华, 丁艳锋, 薛利红, 等.利用叶绿素计(SPAD-502)诊断水稻氮素营养和推荐追肥的研究进展[J].植物营养与肥料学报, 2005(03):412-416.
[10] Schmidt M D, H Lipson.Distilling free-form natural laws from experimentaldata[J].Science, 2009, 324(5923):81-85.
[11] DubcakovaR.Eureqa: software review[J].Genetic Programming and Evolvable Machines, 2011, 12(2): 173-178.
[12] 刘博弈, 王海渝, 龚严, 等.基于天气预报和符号回归算法的参考作物腾发量预测研究[J].中国农村水利水电, 2018(08):22-26.

备注/Memo

备注/Memo:
收稿日期:2018-12-06
基金项目:江苏水利科技项目(2016068)
作者简介:刘文豪(1994—),男,硕士研究生,主要从事节水灌溉与农田生态效应研究工作。
通讯作者:徐俊增(1980—),男,教授,博士,主要从事节水灌溉与农田生态效应研究工作。
更新日期/Last Update: 2019-04-15