Identification of Agaricus Bisporus Disease in Early Stage by Hyperspectral Image
-
摘要: 双孢蘑菇疣孢霉病是由有害疣孢霉菌Mycogone perniciosa引起的、破坏性极强的真菌类病害,且该病害检测困难耗时,往往导致菇房绝收,菇农收益损失严重。早发现、早处理能够有效解决病害带来的经济损失和农药残留超标等质检问题。因此,本研究将能够快速无损检测的高光谱成像技术应用到双孢蘑菇病害早期鉴别。以双孢蘑菇菌Agaricus bisporus子实体为试材,对健康染病双孢蘑菇生长早期子实体样本采集菌盖的全波段(401~1 046 nm)可见/近红外高光谱图像信息,利用多元散射校正(MSC)进行预处理,采用决策树(DT)提取特征波段,对比随机森林(RF)和极限学习机(ELM)两种模型对健康和染病双孢蘑菇鉴别准确度。利用DT选取401.00、951.59、978.09、1 006.59和1 044.90 nm为鉴别病害的特征波段。对比RF和ELM所建模型效果,得到MSC-DT-ELM模型检测效果最优,测试集和预测集总体样本鉴别准确度分别为92.39%和91.32%。结果表明,该模型可以有效提高基于全波段的双孢蘑菇疣孢霉病早期的鉴别准确度,得到基于高光谱成像技术的便捷准确鉴别双孢蘑菇病害早期的模型,同时,为进一步开发双孢蘑菇病害早期的多光谱设备提供了理论依据和方法。Abstract: The wet bubble disease of Agaricus bisporus is a highly destructive fungal disease caused by Mycogone perniciosa, which is difficult to detect and time-consuming, often leading to the failure of mushroom houses and the serious loss of income of the mushroom farmers. Early detection and treatment for this disease can effectively solve the economic loss caused by the diseases and the quality control problems such as excessive pesticide residues. Therefore, in this study, the hyperspectral imaging technology, which could be used for the rapid nondestructive detection, was applied to the early identification of Agaricus bisporus diseases. By using the fruiting body of Agaricus bisporus as the test material, the full-band (401-1046 nm) visible/near-infrared hyperspectral image information of the pileus was collected from the fruiting body samples at the early growth stage of healthy infected Agaricus bisporus. The multiplicative scatter correction (MSC) was used for the preprocessing. The decision tree (DT) was used to extract the characteristic bands. The accuracy of the two models, Random Forest (RF) and Extreme Learning Machine (ELM), for the identification of healthy and infected Agaricus bisporus was compared. By DT, 401.00, 951.59, 978.09, 1006.59 and 1044.90 nm were selected as the characteristic bands for the disease identification. By comparing the effect of RF model and ELM model, the MSC-DT-ELM model had the best detection effect, and the identification accuracy of the test set and the prediction set were 92.39% and 91.32%, respectively. The results showed that the MSC-DT-ELM model could effectively improve the early identification accuracy of the wet bubble disease of Agaricus bisporus based on full band, and obtain a convenient and accurate model for the early identification of Agaricus bisporus disease based on the hyperspectral imaging technology. At the same time, the theoretical basis and method were provided for the further development of multispectral equipment for the rapid and nondestructive identification of Agaricus bisporus diseases in the early stage.
-
-
[1] SIWULSKI M, BUDKA A, RZYMSKI P, et al.Worldwide basket survey of multielemental composition of white button mushroom Agaricus bisporus[J].Chemosphere,2020, 239:124718.(doi: 10.1016/j.chenmosphere.2019.124718)
[2] 孙若兰,肖靓,易有金,等.双孢蘑菇采后贮藏保鲜研究进展[J].食品科学,2021, 42(1):333-340. [3] 张文强,陈青君,张国庆,等.蛋白添加剂对双孢蘑菇农艺性状及品质的影响[J].中国农业科学,2020, 53(10):2091-2100. [4] 张怡,钱忠海,刘琴,等.甲基硫菌灵对季节性栽培双孢蘑菇褐腐病的生物活性及田间防效和安全性评价[J].江苏农业科学,2018, 46(15):81-83. [5] 张春兰,徐济责,李丹,等.有害疣孢霉与不同食用菌的培养关系[J].西北农林科技大学学报(自然科学版),2017, 45(1):112-118. [6] FU Y, WANG X, LI D, et al.Identification of resistance to wet bubble disease and genetic diversity in wild and cultivated strains of Agaricus bisporus[J].International Journal of Molecular Sciences,2016, 17(10):1568.(doi: 10.3390/ijms17101568)
[7] HOLLANDDM,COOKE RC.Nutrient depletion and sporulation in the wet bubble pathogen Mycogone perniciosa[J].Mycological,1991, 3(95):364-369.
[8] 魏峰,侯祥保.双孢蘑菇疣孢霉病的发生及综合防治[J].现代农业科技,2009(2):102. [9] YANG Y, SOSSAH F L, LI Z, et al.Genome-wide identification and analysis of chitinase GH18 gene family in Mycogone perniciosa[J].Frontiers in Microbiology,2021, 11(1):596719.(doi: 10.3389/fmicb.2020.596719)
[10] GEA F J, TELLO J C, NAVARRO M.Efficacy and effects on yield of different fungicides for control of wet bubble disease of mushroom caused by the mycoparasite Mycogone perniciosa[J].Crop Protection,2010, 29(9):1021-1025.
[11] TAO F, YAO H, HRUSKA Z, et al.A novel hyperspectral-based approach for identification of maize kernels infected with diverse Aspergillus flavus fungi[J].Biosystems Engineering,2020, 200:415-430.
[12] GAO Z, KHOT L R, NAIDU R A, et al.Early detection of grapevine leafroll disease in a red-berried wine grape cultivar using hyperspectral imaging[J].Computers and Electronics in Agriculture,2020,179:105807.(doi: 10.1016/j.compag.2020.105807)
[13] BAGNASCO L, ZOTTI M, SITTA N, et al.A PCA-based hyperspectral approach to detect infections by mycophilic fungi on dried porcini mushrooms (boletus edulis and allied species)[J].Talanta,2015, 144:1225-1230.
[14] PARRAG V, FELFÖLDI J, BARANYAI L, et al.Early detection of cobweb disease infection on Agaricus bisporus sporocarps using hyperspectral imaging[J].Acta Alimentaria,2014, 43(S1):107-113.
[15] MARTENS H, NIELSEN J P, ENGELSEN S B.Light ccattering and light absorbance separated by extended multiplicative signal correction.Application to near-Infrared transmission analysis of powder mixtures[J].Analytical Chemistry,2003, 75(3):394-404.
[16] TSAI P, CHEN K, HONG T, et al.Categorizing SHR and WKY rats by chi2 algorithm and decision tree[J].Scientific Reports,2021, 11(1):3463.(doi: 10.1038/s41598-021-82864-3)
[17] BELGIU M, DRÂGUŢ L.Random forest in remote sensing:A review of applications and future directions[J].ISPRS Journal of Photogrammetry and Remote Sensing,2016, 114:24-31.
[18] HUANG G B, ZHU Q Y, SIEW C K.Extreme learning machine:Theory and applications[J].Neurocomputing,2006, 70(1/3):489-501.
[19] DING S, XU X, NIE R.Extreme learning machine and its applications[J].Neural Computing and Applications,2014, 25(3-4):549-556.
[20] 李鸿强.基于高光谱分析的蔬菜品质检测方法研究[D].北京:中国农业大学,2019. 2021年第4期余松金,黄佳维,李和平,等.不同基质对石斛兰分蘖率和新芽高度的影响[J].福建农业科技,2021,52(4):23-26. -
期刊类型引用(2)
1. 邹金平,魏萱. 基于显微高光谱图像的双孢蘑菇疣孢霉菌厚垣孢子目标检测. 福建农业科技. 2022(02): 1-8 . 本站查看
2. 杨东,王舒卉,吴建华,姜俊伊,宋凯,石天玉. 玉米籽粒霉变等级高光谱图像检测方法研究. 中国粮油学报. 2022(11): 46-53 . 百度学术
其他类型引用(0)
计量
- 文章访问数: 380
- HTML全文浏览量: 0
- PDF下载量: 186
- 被引次数: 2