基于支持向量机的玉米苗期田间杂草光谱识别

doi:10.3964/j.issn.1000-0593(2009)07-1906-05

Abstract
References
Related Citation (15)

Download: PDF (1334 KB)
Export: BibTeX | EndNote (RIS)

Abstract A handheld FieldSpec^® 3 Spectroradiometer manufactured by ASD Incorporated Company in USA was used to measure the spectroscopic data of canopies of seedling corns, Dchinochloa crasgalli, and Echinochloa crusgalli weeds within the 350-2 500 nm wavelength range in the field. Each canopy was measured five times continuously. The five original spectroscopic data were averaged over the whole wavelength range in order to eliminate random noise. Then the averaged original data were converted into reflectance data, and the unsmooth parts of reflectance spectral curves with large noise were removed. The effective wavelength range for spectral data process was selected as 350-1 300 and 1 400-1 800 nm. Support vector machine (SVM) was chosen as a method of pattern recognition in this paper. SVM has the advantages of solving the problem of small sample size, being able to reach a global optimization, minimization of structure risk, and having higher generalization capability. Two classes of classifier SVM models were built up respectively using “linear”, “polynomial”, “RBF”(radial basis function), and “mlp (multilayer perception)” kernels. Comparison of different kernel functions for SVM shows that higher precision can be obtained by using “polynomial” kernel function with 3 orders. The accuracy can be above 80%, but the SV ratio is relatively low. On the basis of two-class classification model, taking use of voting procedure, a model based on one-against-one-algorithm multi-class classification SVM was set up. The accuracy reaches 80%. Although the recognition accuracy of the model based on SVM algorithm is not above 90%, the authors still think that the research on weeds recognition using spectrum technology combining SVM method discussed in this paper is tremendously significant. Because the data used in this study were measured over plant canopies outdoor in the field, the measurement is affected by illumination intensity, soil background, atmosphere temperature and instrument accuracy. This method proposes a kind of research ideology and application foundation for weeds recognition in the field.

Key words：Spectral;Pattern recognition;Support vector machines (SVM);Corn;Weeds

Received: 2008-12-02 Accepted: 2009-03-06

ZTFLH:

O657.3

Corresponding Authors: HE Xiong-kui E-mail: xiongkui@cau.edu.cn

Cite this article:

DENG Wei,ZHANG Lu-da,HE Xiong-kui, et al. SVM-Based Spectral Recognition of Corn and Weeds at Seedling Stage in Fields [J]. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2009, 29(07): 1906-1910.

URL:

https://www.gpxygpfx.com/EN/10.3964/j.issn.1000-0593(2009)07-1906-05 OR https://www.gpxygpfx.com/EN/Y2009/V29/I07/1906

[1] QIANG Sheng, NI Han-wen, JIN Yin-gen, et al(强胜, 倪汉文, 金银根, 等). Weed Science(杂草学). Beijing: China Agriculture Press(北京: 中国农业出版社), 2007. 29.
[2] GUO Zhen-sheng(郭振升). J. of Henan Agric.(河南农业), 2003, (9): 32.
[3] LI Xiang-ju(李香菊). The Weed Chemical Control in Corn and Coarse Cereals Fields(玉米及杂粮田杂草化学防除). Beijing: Chemical Industry Press(北京: 化学工业出版社), 2003.
[4] HE Xiong-kui(何雄奎). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2004, 20(1): 13.
[5] DENG Wei, DING Wei-min(邓巍，丁为民). Transactions of the Chinese Society of Agricaltural Machinery(农业机械学报), 2008, 39(6): 77.
[6] Giles D K, Andersen P G, Nilars M. Transactions of the ASAE, 2002, 45(3): 539.
[7] Geol P K, Prasher S O, Patel R M, et al. Transactions of the ASAE, 2002, 45(2): 443.
[8] MAO Wen-hua, WANG Yi-ming, ZHANG Xiao-chao, et al(毛文华, 王一鸣, 张小超, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2004, 20(5): 43.
[9] Wang N, Zhang N, Dowell F E, et al. Transactions of the ASAE, 2001, 44(2): 409.
[10] Slaughter D C, Lanini W T, Giles D K. Transactions of the ASAE, 2004, 47(6): 1907.
[11] Brown R B, et al. Transactions of the ASAE, 1994, 37(1): 297.
[12] Vrindts E, et al. Proceedings of the 2<sup>nd</sup> European Conference on Precision Agriculture, 1999, 257.
[13] MAO Wen-hua, WANG Yue-qin, WANG Yi-ming, et al(毛文华, 王月青, 王一鸣, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2005, 25(6): 984.
[14] ZHU Deng-sheng, PAN Jia-zhi, HE Yong(朱登胜, 潘家志, 何勇). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2008, 28(5): 1102.
[15] LI Fang-fang, ZHAO Ying-kai, YAN Xin(李方方, 赵英凯, 颜昕). Computer Application(计算机应用), 2006, 26(12): 358.
[16] ZHANG Yan, FENG Shi-qiang, PAN Da-zhi, et al(张炎, 冯世强, 潘大志, 等). Application of Statistics and Management(数理统计与管理), 2008, 27(2): 211.
[17] BAI Peng, JI Juan-zao, ZHANG Fa-qi, et al(白鹏, 冀捐灶, 张发启, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2008, 28(10): 2278.
[18] Vapnik V. The Nature of Statistical Learning Theory. New York: Springer-Verlag, 1995.
[19] LI Ying, BAI Ben-du, JIAO Li-cheng(李映, 白本督, 焦李成). Systems Engineering and Electronics(系统工程与电子技术), 2001, 23(9): 25.
[20] Chapelle O. Vapnik V. Machine Learning, 2002, 46: 131.
[21] YANG Shu-ying(杨淑莹). Pattern Recognition & Intelligent Computation-Matlab Tech.(模式识别与智能计算·Matlab技术). Beijing: Publishing House of Electronics Industry(北京: 电子工业出版社), 2008. 135.