基于高光谱的核桃叶片铁元素含量反演模型研究

doi:10.6048/j.issn.1001-4330.2018.07.011

摘要/Abstract

摘要： 【目的】铁元素是评价植物长势的重要指标,监测营养状况是果树栽培管理过程中的主要指标之一。分析核桃叶片Fe素含量与叶片光谱之间的关系,监测核桃活体叶片Fe素含量,为核桃叶片Fe含量快速诊断提供理论依据和技术支撑。【方法】对阿克苏红旗坡地区8个核桃生产园进行随机采样,测定其叶片光谱反射率及Fe素含量,对经过数据变换后的光谱采用逐步线性回归和主成分回归两种方法构建Fe素含量估测模型。用随机分组法选取检验样本并进行精度评价。【结果】经过数据转换的光谱与Fe素含量的相关性明显高于原始光谱,其中以对数二阶光谱725~730 nm、1 140、1 141、1 443、1 444和1 445 nm作为敏感波段构建的主成分回归模型精度最佳,构建模型为y=558.648+113.024 PC₁³-31.514 PC₁²+71.519 PC₁,R²为0.870,RMSE为0.304,RPO为2.39。【结论】光谱技术在开展阿克苏红旗坡地区核桃叶片Fe素含量估算方面具有一定的应用潜力。

关键词: 核桃; 光谱反射率; 主成分回归; 逐步回归

Abstract: 【Objective】 Fe element is an important index to evaluate plant growth, and monitoring nutritional status is one of the main contents in the process of fruit tree cultivation and management. This project aims to realize monitoring of foliar Fe content in Juglans regia L.in vivo by analyzing the relationship between foliar Fe content in Juglans regia L. and foliar spectra in the hope of providing theoretical basis and technical support for rapid diagnosis of Fe content.【Method】Eight Juglans regia L. production garden were selected in Aksu Hongqipo area for random sampling, the spectral reflectance and Fe content of the foliar were measured. The method of stepwise linear regression and principal component regression was used to construct the estimation model of Fe content. Random grouping method was used to select test samples and evaluate the accuracy.【Result】The correlation between the content of Fe in Juglans regia L. foliar and spectra transformed by differential, logarithm and so on was significantly higher than that of the original spectrum. In these data conversions, the highest accuracy of the model was principal component regression model, which was based on the logarithmic second-order spectra of 725-730 nm, 1,140 nm, 1,141 nm, 1,443 nm, 1,444 nm and 1,445 nm. So the optimal model for monitoring foliar Fe element of Juglans regia L. was y=558.648+113.024 PC₁³-31.514 PC₁² +71.519 PC₁, R² was 0.870, RMSE was 0.304, RPO was 2.39.【Conclusion】Spectral technique has a certain application potential in estimating the Fe content of Juglans regia L. in Hongqipo area of Aksu.

Key words: Juglans regia L.; spectral reflectance; principal component regression; gradual regression

中图分类号:

S664.1

瞿余红, 王振锡, 丁雅, 刘玉霞, 董淼, 李园. 基于高光谱的核桃叶片铁元素含量反演模型研究[J]. 新疆农业科学, 2018, 55(7): 1264-1273.

QU Yu-hong, WANG Zhen-xi, DING Ya, LIU Yu-xia, DONG Miao, LI Yuan. Research on the Inversion Model for Predicting Fe Content in Juglans regia L. Leaves Based on Hyperspectral Method[J]. Xinjiang Agricultural Sciences, 2018, 55(7): 1264-1273.

参考文献

[1] Broge, N. H., & Mortensen, J. V. (2002). Deriving green crop area index and canopy chlorophyll density of winter wheat from spectral reflectance data. Remote Sensing of Environment, 81(1): 45-57.
[2] 王植, 周连第, 李红, 等. 桃树叶片氮素含量的高光谱遥感监测[J]. 中国农学通报, 2011, 27(4): 85-90.
   WANG Zhi, ZHOU Liang-di, LI Hong, et al. (2011). Predicting nitrogen concentrations in fresh peach leaf hyper spectral remote sensing [J]. Chinese Agricultural Science Bulletin, 27(4): 85-90. (in Chinese)
[3] 罗万杰. 新疆特色林果业发展现状、存在问题及对策[J]. 新疆农垦科技, 2012, (8): 16-18.
   LUO Wan-jie. (2012). The development status, problems and countermeasures of xinjiang characteristic forest fruit industry [J]. Xinjiang Farm Research of Science and Technology, (8): 16-18. (in Chinese)
[4] 邢东兴, 常庆瑞. 基于光谱分析的果树叶片微量元素含量估测研究--以红富士苹果树为例[J]. 西北农林科技大学学报, 2008. 11(36): 017.
   XING Dong-xing, CHANG Qing-rui. (2008).Research on predicting the Fe, Mn, Cu, Zn contents in fruit trees' fresh leaves by spectral analysis-Red Fuji apple tree as an example [J]. Journal of Northwest A & F University, 11(36): 017. (in Chinese)
[5] 崔澂. 微量元素在植物中的生理作用及其在农业实践上的意义[J]. 科学通报, 1957, (17): 524-529.
   CUI Ce. (1957). The Physiological role of trace elements in plants and its significance in agricultural practice [J]. Science Bulletin, (17): 524-529. (in Chinese)
[6] 冯振铭, 石雪. 果树园艺技术与果树营养诊断策略之我见[J]. 农业与技术, 2016, (2): 219.
   FENG Zhen-ming, SHI Xue. (2016).My Opinion on fruit tree horticulture technique and nutritional diagnosis of fruit tree [J]. Agriculture and Technology, (2): 219. (in Chinese)
[7] 王晓星. 夏玉米冠层光谱特征及其生理生态参量的高光谱估算模型[D]. 杨凌: 西北农林科技大学硕士论文, 2015.
   WANG Xiao-xing. (2015). Hyperspectral characteristic and estimating models about physiological ecological parameters of corn in canopy level [D]. Master Dissertation. Northwest A & F University, Yangling. (in Chinese)
[8] 王强, 易秋香, 包安明, 等. 基于高光谱反射率的棉花冠层叶绿素密度估算[J]. 农业工程学报, 2012, 28(15): 125-132.
   WANG Qiang, YI Qiu-xiang, BAO An-ming, et al. (2012).Estimating chlorophyll density of cotton canopy by hyperspectral reflectance [J] Transactions of the Chinese Society of Agricultural Engineering, 28(15): 125-132. (in Chinese)
[9] 黎瑞君, 聂克艳, 彭志良, 等. 辣椒盛果期的冠层反射光谱特征[J]. 贵州农业科学, 2014, 42(12): 230-233.
   LI Rui-jun, NIE Ke-yan, PENG Zhi-liang, et al. (2014).Canopy spectral reflectance characteristics of pepper during full bearing period [J]. Guizhou Agricultural Sciences, 42(12): 230-233.
[10] 李颖, 薛利红, 潘复燕, 等. 氮磷互作对水稻冠层光谱的影响及其PNN识别[J]. 中国农业科学, 2014, 47(14): 2 742-2 750.
   LI Ying, XUE Li-hong, PAN Fu-yan, et al. (2014).Effects of interaction of N and P on rice canopy spectral reflectance and its PNN identification [J]. Agricultural Sciences in China, 47(14): 2,742-2,750. (in Chinese)
[11] 尚艳. 不同氮水平下小麦冠层光谱特征及其与农学参数关系研究[D]. 杨凌: 西北农林科技大学硕士论文, 2015.
   SHANG Yan. (2015). Wheat canopy spectral features and its research relationship with agronomy parameter under different nitrogen levels[D]. Master Dissertation. Northwest A & F University, Yangling. (in Chinese)
[12] 胡珍珠, 潘存德, 王世伟, 等. 轮台白杏叶片氮磷钾含量光谱估算模型[J]. 新疆农业科学, 2013, 50(2): 238-248.
   HU Zhen-zhu, PAN Cun-de, WANG Shi-wei, et al. (2013). Model for estimating foliar NPK content ofarmeniaca vulgaris 'Luntaibaixing' using spectral reflectance [J]. Xinjiang Agricultural Sciences, 50(2): 238-248.
[13] 冀荣华, 郑立华, 邓小蕾, 等. 基于反射光谱的苹果叶片叶绿素和含水率预测模型[J]. 农业机械学报, 2014, 45(8): 269-275.
   JI Rong-hua, ZHENG Li-hua, DENG Xiao-lei, et al. (2014).Prediction model of chlorophyll and water cut in apple leaves based on reflectance spectrum [J]. Transactions of the Chinese Society for Agricultural Machinery, 45(8): 269-275. (in Chinese)
[14] 邓小玲, 郑建宝, 梅慧兰, 等. 基于高光谱成像技术的柑橘黄龙病病情诊断及分类[J]. 西北农林科技大学学报(自然科学版), 2013, 41(7): 99-105.
   DENG Xiao-ling, ZHENG Jian-bao, MEI Hui-lan, et al. (2013).Identification and classification of citrus Huanglongbing disease based on hyperspectral imaging [J] Journal of Northwest A & F University (Natural Science Edition) , 41(7): 99-105. (in Chinese)
[15] 徐爽. 基于高光谱图像技术的红枣品质无损检测研究[D]. 银川: 宁夏大学硕士论文, 2013.
   XU Shuang. (2013). Study on nondestructive detection of jujube quality based on hyperspectral imaging technology [D]. Master Dissertation. Ningxia University, Yingchuan. (in Chinese)
[16] Yoder, B. J., Pettigrewcrosby, R. E. (1995). Predicting nitrogen and chlorophyll content and concentrations from reflectance spectra (400-2500nm) at leaf and canopy scales. Remote Sensing of Environment, 53(3): 199-211.
[17] 胡珍珠, 潘存德, 王世伟, 等. 轮台白杏叶片铁锰浓度光谱估算模型[J]. 光谱学与光谱分析, 2014, 34(9): 2 485-2 489.
   HU Zhen-zhu, PAN Cun-de, WANG Shi-wei, et al. (2014). Model for estimating foliar Fe、Mn content ofarmeniaca vulgaris 'Luntaibaixing' using spectral reflectance [J]. Spectroscopy and Spectral Analysis, 34(9): 2,485-2,489. (in Chinese)
[18] 马超飞, 马建文, 韩秀珍. 微量元素在植物光谱中的响应机理研究[J]. 遥感学报, 2001, 5(5): 334-339.
   MA Chao-fei, MA Jian-wen, HAN Xiu-zhen. (2001). Response Mechanism of Trace Elements in Plant Spectra [J]. Journal of Remote Sensing, 5(5): 334-339.
[19] 郑咏梅, 张铁强, 张军, 等. 平滑、导数、基线校正对近红外光谱PLS定量分析的影响研究[J]. 光谱学与光谱分析, 2004, (12): 1 546-1 548.
   ZHENG Yong-mei, ZHANG Tie-qiang, ZHANG Jun, et al. (2004). The effects of smooth, derivative and baseline correction on the quantitative analysis of near infrared spectrum PLS [J]. Spectroscopy and Spectral Analysis, (12): 1,546-1,548. (in Chinese)
[20] 田高友, 袁洪福, 褚小立, 等. 结合小波变换与微分法改善近红外光谱分析精度[J]. 光谱学与光谱分析, 2005, (4): 516-520.
   TIAN Gao-you, YUAN Hong-fu, CHU Xiao-li, et al. (2005). The precision of near-infrared spectroscopy is improved by combining wavelet transform with differential method [J]. Spectroscopy and Spectral Analysis, (4): 516-520. (in Chinese)
[21] 张飞, 丁建丽, 等. 渭干河-库车河三角洲绿洲盐渍化地地物光谱数据分析[J]. 光谱学与光谱分析, 2008, 28(12): 2 921-2 926.
   ZHANG Fei, DING Jian-li, Tashpolat Tiyip, et al. (2008). Analysis of the spectral data of salinized land in the weigan river - kuchaihe delta oasis [J]. Spectroscopy and Spectral Analysis, 28(12): 2,921-2,926. (in Chinese)
[22] 张东, 张飞,等. 分数阶微分在盐渍土高光谱数据预处理中的应用[J]. 农业工程学报, 2014, 30(24): 151-160..
   ZHANG Dong, Tashpolat Tiyip, ZHANG Fei, et al. (2014). Application of fractional differential in preprocessing hyperspectral data of saline soil [J]. Transactions of the Chinese Society of Agricultural Engineering, 30(24): 151-160. (in Chinese)
[23] 王磊, 白由路, 卢艳丽, 等. 光谱数据变换对玉米氮素含量反演精度的影响[J]. 遥感技术与应用, 2011, 26(2): 220-225.
   WANG Lei, BAI You-lu, LU Yan-li, et al. (2011). The effect of spectral data transformation on the accuracy of nitrogen content in maize [J]. Remote Sensing Technology and Application, 26(2): 220-225. (in Chinese)
[24] 王乾龙. 基于土壤可见-近红外光谱库的土壤全氮预测建模研究[D]. 杭州: 浙江大学硕士论文, 2015.
   WANG Qian-long. (2015). Predicting soil total nitrogen based on the soil visible and near-infrared spectral library [D]. Master Dissertation. Zhejiang University, Hangzhou. (in Chinese)
[25] 田烨. 基于实验光谱的土壤锰含量估算研究[D]. 南京: 南京信息工程大学硕士论文, 2014.
   TIAN Ye. (2014). Study on estimation of soil manganese content based on experiment spectra [D]. Master Dissertation. Nanjing University of Information Science & Technology, Nanjing. (in Chinese)
[26] Barnes, R. J., Baxter, S. J., Lark, R. M. (2007). Spatial covariation of Azotobacter, abundance and soil properties: A case study using the wavelet transform.Soil Biology & Biochemistry, 39(1): 295-310.
[27] 余天霞, 罗红霞, 邹扬庆, 等. 北碚447锦橙叶片农学参数与其反射光谱的相关性研究[J]. 西南师范大学学报(自然科学版), 2014, 39(5): 33-37.
   YU Tian-xia, LUO Hong-xia, ZHOU Yang-qing, et al. (2014). On correlation between leaf spectral reflectance and agricultural parameter of Beibei 447 Jincheng orange leaves [J]. Journal of Southwest China Normal University (Natural Science Ed.) , 39(5): 33-37. (in Chinese)
[28] 祁琼. 基于高光谱数据的苔草营养成分反演方法研究[J]. 地理空间信息, 2017, 15(10): 90-93,11.
   QI Qiong. (2017). A study on the inversion method of liverwort nutrient components based on hyperspectral data [J]. Geospatial Information, 15(10): 90-93,11. (in Chinese)
[29] 浦瑞良, 宫鹏. 高光谱遥感及其应用[M]. 北京: 高等教育出版社, 2003.
   PU Rui-liang, GONG Peng. Hyperspectral remote sensing and its application [M]. Beijing: Higher Education Press, 2003.
[30] Lee K S, Cohen W B, Kennedy R E, et al. (2004). Hyperspectral versus multispectral data for estimating leaf area index in four different biomes. Remote Sens. Environ, (91): 508-520.
[31] 黄文珂. 多元回归建模过程中共线性的诊断与解决方法[D]. 哈尔滨: 哈尔滨工业大学硕士论文, 2012.
   HUANG Wen-ke. (2012). The diagnosis and process solutions in multicollinearity of multiple regression model[D]. Master Dissertation. Harbin Institute of Technology, Harbin. (in Chinese)
[32] 胡珍珠, 潘存德, 潘鑫, 等. 基于光谱水分指数的核桃叶片含水量估算模型[J]. 林业科学, 2016, 52(12): 39-49.
   HU Zhen-zhu, PAN Cuo-de, PAN Xin, et al. Estimation models for water content of walnut leaves based on spectral moisture index[J]. Scientia Silvae Sinicae, 2016, 52(12): 39-49.
[33] 邵咏妮. 水稻生长生理特征信息快速无损获取技术的研究[D]. 杭州: 浙江大学博士论文, 2010.
   SHAO Yong-ni. (2010). Research on non-destructive and rapid acquisition technique for rice physiological characteristics and growth information [D]. PhD Dissertation. Zhejiang University, Hangzhou. (in Chinese)
[34] 胡珍珠. '温185'核桃叶片营养元素含量与坚果品质指标光谱反演[D]. 乌鲁木齐: 新疆农业大学, 2016.
   HU Zhen-zhu. (2016). Spectrum inversion of foliar mineral nutrient concentrations and nut quality indexes for Juglans regia'Wen 185' [D]. Urumqi: Xinjiang Agricultural University.