ru
Back
DOI: 10.17586/1023-5086-2018-85-07-69-75
Early identification of Curvularia Lunata based on hyperspectral imaging
Full text «Opticheskii Zhurnal»
Full text on elibrary.ru
Publication in Journal of Optical Technology
Jing Xu, Teng Miao, Yuncheng Zhou, Hanbing Deng, Ping Song, Yubo Zhang, Kai Song Early identification of Curvularia Lunata based on hyperspectral imaging (Ранняя идентификация Curvularia Lunata с использованием гиперспектральных изображений) [на англ. яз.] // Оптический журнал. 2018. Т. 85. № 7. С. 69–75. http://doi.org/10.17586/1023-5086-2018-85-07-69-75
Jing Xu, Teng Miao, Yuncheng Zhou, Hanbing Deng, Ping Song, Yubo Zhang, Kai Song Early identification of Curvularia Lunata based on hyperspectral imaging (Ранняя идентификация Curvularia Lunata с использованием гиперспектральных изображений) [in English] // Opticheskii Zhurnal. 2018. V. 85. № 7. P. 69–75. http://doi.org/10.17586/1023-5086-2018-85-07-69-75
Jing Xu, Teng Miao, Yuncheng Zhou, Hanbing Deng, Ping Song, Yubo Zhang, and Kai Song, "Early identification of Curvularia lunata based on hyperspectral imaging," Journal of Optical Technology. 85(7), 432-436 (2018). https://doi.org/10.1364/JOT.85.000432
With the popularizing of close planting and high fertilizer technology, the occurrence of Curvularia Lunata has been increasing year after year, has caused a very serious harm. The early identification method of the disease is of great significance to reduce the environmental pollution and improve the quality of crops. In order to identify the disease quickly and non-damage, we studied on maize leaves, then propose an early identification of Curvularia Lunata method based on hyperspectral imaging technology. The experiment leaves are in vitro inoculation, at 11 time's points after inoculation (2, 4, 6, 8, 12, 24, 32, 40, 48, 60, 72 hours respectively), we collect hyperspectral images in the visible and near infrared bands of normal leaves and inoculation disease leaves respectively by hyperspectral imaging system. By comparative analysis between two types of leaves, small chlorotic spots are been found after inoculation with 48 hours, spectral information has obvious difference. Based on mixed distance method, we find that the best bands of identification Curvularia Lunata are 465.1,550.7, 681.4 nm. We use these bands as the input quantity and build back-propagation neural network detection model. We make experiments on samples (160 samples for each time point above mentioned) by the model, the accuracy rate of identification inoculation leaves is above 97.5%. The result show that, the BP neural network detection model with characteristic bands can identification Curvularia Lunata quickly and non-destructive. It provides a new way for early detection of maize diseases.
hyperspectral imaging technology, Curvularia Lunata, characteristic band, early identification, non-destruction detection
Acknowledgements:This research was financially supported by Natural Science Foundation of China (31501217, 31601218, 31701318).
OCIS codes: 100.4145, 150.1135, 170.0110, 300.6280, 300.6550, 000.5490
References:1. Guo Q.H. The development and evolution of the major Maize producing areas in China // J. Maize. Sci. 2010. V. 18(1). P. 139–145.
2. Fu J.F., Li H.C., Bai Y.J. Spread gradient model of leaf spot disease (Curvularia Lunata) in Maize // Acta. Phytopathologica Sinica. 2003. V. 33(5). P. 456–461.
3. Li J.T., Mo S.X. Effects of two-plants orientation cultivation pattern on the maize yield and occurrence and epidemic of curvularia leaf spot // J. Maize. Sci. 2015. V. 23(2). P. 137–140, 146.
4. Li J.B., Rao X.Q., Ying Y.B. Advance on application of hyperspectral imaging to nondestructive detection of agricultural products external quality // Spectrosc. Spect. Anal. 2011. V. 31(8). P. 2021–2026.
5. Ma B.X., Ying Y.B., Rao X.Q. Advance in nondestructive detection of fruit internal quality based on hyperspectral imaging // Spectrosc. Spect. Anal. 2009. V. 29(6). P. 1611–1615.
6. Chen J.J., Peng Y.K., Li Y.Y. Rapid detection of vegetable pesticide residue based on hyperspectral fluorescence imaging technology // Trans. CSAE. 2010. V. 26(14). P. 1–5.
7. Huang W.Q., Chen L.P., Li J.B. Effective wavelengths determination for detection of slight bruises on apples based on hyperspectral imaging // Trans. CSAE. 2013. V. 29(1). P. 272–277.
8. Tian Y.W., Cheng Y., Wang X.Q. Feature vectors determination for pest detection on apples based on hyperspectral imaging // Trans. CSAE. 2014. V. 30(12). P. 132–138.
9. Lorente D., Aleixos N., Gómez-Sanchis J. Recent advances and applications of hyperspectral imagine for fruit and vegetable quality assessment // Food. Bioprocess. Tech. 2012. V. 5(4). P. 1121–1142.
10. Shan J.J., Wu J.H., Chen J.J. Rapid nondestructive detection of apple quality attributes using hyperspectral scattering images // Spectrosc. Spect. Anal. 2010. V. 30(10). P. 2729–2733.
11. Zhang C., Liu F., Kong W.W. Fast identification of watermelon seed variety using near infrared hyperspectral imaging technology // Trans. CSAE. 2013. V. 29(20). P. 270–277.
12. Zhu Q.B., Feng Z.L., Huang M. Maize seed classification based on image entropy using hyperspectral imaging technology // Trans. CSAE. 2012. V. 28(23). P. 271–276.
13. Qiao J., Ngadi M., Wang N. Pork quality and marbling level assessment using a hyperspectral imaging system // J. Food. Eng. 2007. V. 83(1). P. 10–16.
14. Zhang L.L., Li Y.Y., Peng Y.K. Determination of pork freshness attributes by hyperspectral imaging technique // Trans. CSAE. 2012. V. 28(7). P. 254–259.
15. Kamruzzaman M., Makino Y., Oshita S. Rapid and non-destructive detection of chicken adulteration in minced beef using visible near-infrared hyperspectral imaging and machine learning // J. Food. Eng. 2016. V. 170(2). P. 8–15.
16. Lawrence K.C., Yoon S.C., Heitschmidt G.W. Imaging system with modified-pressure chamber for crack detection in shell eggs // Sensing and Instrumentation for Food Quality and Safety. 2008. V. 2(3). P. 116–122.
17. Moshou D., Bravo C., West J. Automatic detection of ‘yellow rust’ in wheat using reflectance measurements and neural network // Comput. Electron. Agr. 2004. V. 44(3). P. 173–188.
18. Moshou D., Bravo C., Oberti R. Plant disease detection based on data fusion of hyper-spectral and multi-spectral fluorescence imaging using Kohonen maps // Real-Time Imaging. 2005. V. 11(2). P. 75–83.
19. Feng W., Wang X.Y., Song X. Hyperspectral estimation of canopy chlorophyll density in winter wheat under stress of powdery mildew // Trans. CSAE. 2013. V. 29(13). P. 114–123.
20. Zheng Z.X., Qi L., Ma X. Grading method of rice leaf blast using hyperspectral imaging technology // Trans. CSAE. 2013. V. 29(19). P. 138–144.
21. Tian Y.W., Li T.L., Zhang L. Diagnosis method of cucumber disease with hyperspectral imaging in greenhouse // Trans. CSAE. 2010. V. 26(5). P. 202–206.
22. Li B., Liu Z.Y., Huang J.F. Hyperspectral identification of rice diseases and pests based on principal component analysis and probabilistic neural network // Trans. CSAE. 2009. V. 25(9). P. 43–147.
23. Yuan Y., Wang W., Chu X. Detection of corn aflatoxin based on hyperspectral imaging technology and factor discriminant analysis // J. Chinese Cereals and Oils Association. 2014. V. 29(12). P. 107–110.