VideometerLab 4多光譜種子表型成像系統(tǒng)是丹麥理工大學(xué)與丹麥Videometer公司開發(fā),是用于種子研究先進(jìn)的多光譜表型成像設(shè)備,典型客戶為ISTA國(guó)際種子檢驗(yàn)協(xié)會(huì)、ESTA歐洲種子檢驗(yàn)協(xié)會(huì)、John Innes Centre、LGC化學(xué)家集團(tuán)、奧胡斯大學(xué)等等,利用該系統(tǒng)發(fā)表的文章已經(jīng)超過(guò)300篇。
Videometer種子表型表型成像系統(tǒng)可測(cè)量種子如尺寸、顏色、形狀等,間接測(cè)定種子參數(shù)如種子純度、發(fā)芽百分比、發(fā)芽率、種子健康度、種子成熟度、中壽命等。種子活力綜合種子活力是種子發(fā)芽和出苗率、幼苗生長(zhǎng)的潛勢(shì)、植株抗逆能力和生產(chǎn)潛力的總和(發(fā)芽和出苗期間的活性水平與行為),是種子品質(zhì)的重要指標(biāo),具體包括吸漲后旺盛的代謝強(qiáng)度、出苗能力、抗逆性、發(fā)芽速度及同步性、幼苗發(fā)育與產(chǎn)量潛力。種子活力是植物的重要表型特征,傳統(tǒng)檢測(cè)方法包括低溫測(cè)試、高溫加速衰老測(cè)試、幼苗生長(zhǎng)測(cè)定等。
該系統(tǒng)也可以對(duì)細(xì)菌、蟲卵、真菌等進(jìn)行高通量成像測(cè)量,進(jìn)行病理學(xué)、毒理學(xué)或其它研究。對(duì)于擬南芥等冠層平展的植物,可以進(jìn)行自動(dòng)的葉片計(jì)數(shù)等。
摘要
玉米不可避免地受到玉米赤霉烯酮(zearalenone,ZEN)的污染,對(duì)人類造成嚴(yán)重危害。本研究采用多光譜成像(MSI)技術(shù)結(jié)合不同的機(jī)器學(xué)習(xí)方法檢測(cè)玉米中的ZEN含量。利用遺傳算法和反向傳播神經(jīng)網(wǎng)絡(luò)(GA-BPNN)可以選擇與玉米中ZEN濃度最相關(guān)的波長(zhǎng)。結(jié)果表明,GA-BPNN方法可以檢測(cè)出ZEN污染水平,準(zhǔn)確率為93.33%。此外,GA-BPNN算法是定量預(yù)測(cè)ZEN污染含量的最佳方法,預(yù)測(cè)集的相關(guān)系數(shù)(R p)、均方根誤差(RMSEP)、殘差預(yù)測(cè)偏差(RPD)和偏差分別達(dá)到0.95、3.66μg/kg、5.39和1.55μg/kg?梢缘贸鼋Y(jié)論,多光譜成像與機(jī)器學(xué)習(xí)相結(jié)合適用于玉米禪宗含量的快速測(cè)量。
關(guān)鍵詞:多光譜成像、無(wú)損檢測(cè)、玉米赤霉烯酮、機(jī)器學(xué)習(xí)方法
Application of multispectral imaging combined with machine learning methods for rapid and non-destructive detection ofzearalenone (ZEN) in maize
Abstract
Maize is inevitably contaminated by zearalenone (ZEN) that will cause serious harm to human beings. In this study, multispectral imaging (MSI) technology combined with different machine learning were used to detect ZEN content in maize. The wavelengths that were most related to ZEN concentration in maize could be selected by genetic algorithm with a back-propagation neural network (GA-BPNN). Our results showed that ZEN contamination level could be detected with the accuracy of 93.33% by GA-BPNN method. In addition, for quantitative prediction of ZEN contamination content GA-BPNN algorithm was the best method with the correlation coefficient (R p ), the root means square error (RMSEP), residual predictive deviation (RPD) and bias achieved to 0.95, 3.66 μg/kg, 5.39 and 1.55 μg/kg, respectively in prediction set. It can be concluded that multispectral imaging combined with machine learning was applicable for rapid measurement of ZEN content in maize.
Key words: Multispectral imaging, Non-destructive detection, Zearalenone in maize,Machine learning method