国产日韩欧美一区二区三区三州_亚洲少妇熟女av_久久久久亚洲av国产精品_波多野结衣网站一区二区_亚洲欧美色片在线91_国产亚洲精品精品国产优播av_日本一区二区三区波多野结衣 _久久国产av不卡

?

基于圖像和光譜信息融合的紅茶萎凋程度量化判別

2017-01-09 06:47:30寧井銘孫京京朱小元李姝寰張正竹黃財(cái)旺
關(guān)鍵詞:兒茶素紅茶特征值

寧井銘,孫京京,朱小元,李姝寰,張正竹,黃財(cái)旺

?

基于圖像和光譜信息融合的紅茶萎凋程度量化判別

寧井銘1,孫京京1,朱小元1,李姝寰1,張正竹1,黃財(cái)旺2

(1. 安徽農(nóng)業(yè)大學(xué)茶樹生物學(xué)與資源利用國(guó)家重點(diǎn)實(shí)驗(yàn)室,合肥 230036; 2. 安徽祁門金東茶廠,祁門 245600)

為了實(shí)現(xiàn)對(duì)紅茶萎凋程度量化判別,該研究提出了一種將圖像和光譜信息融合后分別與線性判別分析法和偏最小二乘法結(jié)合的技術(shù),進(jìn)行工夫紅茶萎凋程度定性判別及兒茶素與氨基酸比值定量預(yù)測(cè)研究。通過對(duì)圖像進(jìn)行主成分分析,篩選出5個(gè)特征波長(zhǎng)和對(duì)應(yīng)的光譜特征值,基于灰度共生矩陣提取5個(gè)特征波長(zhǎng)圖像的紋理特征值,并采用連續(xù)投影算法優(yōu)選出14個(gè)紋理特征值,然后分別以光譜和紋理特征值融合數(shù)據(jù)建立紅茶萎凋程度的線性判別模型和兒茶素與氨基酸比值的偏最小二乘預(yù)測(cè)模型。結(jié)果表明:采用所研究的方法和建立的模型對(duì)工夫紅茶萎凋程度判別準(zhǔn)確率達(dá)到94.64%,兒茶素與氨基酸比值預(yù)測(cè)相關(guān)系數(shù)為0.8765,預(yù)測(cè)均方根誤差為0.434,預(yù)測(cè)結(jié)果較好。證明應(yīng)用這兩種方法能實(shí)現(xiàn)對(duì)紅茶萎凋程度量化判別。

數(shù)據(jù)融合;判別分析方法;圖像分析;偏最小二乘法;紅茶;萎凋;兒茶素與氨基酸比值

0 引 言

紅茶加工分為初制和精制2個(gè)過程,初制過程包括萎凋、揉捻(揉切)、發(fā)酵和干燥4道工序,其中萎凋是紅茶初制的首要工序和品質(zhì)形成的基礎(chǔ)。茶鮮葉在萎凋過程中,既有物理變化,也有化學(xué)變化。物理變化表現(xiàn)為葉質(zhì)變軟,葉色從翠綠變?yōu)榘稻G;同時(shí),隨著鮮葉含水量的降低,內(nèi)部化學(xué)成分也發(fā)生一系列變化,萎凋適度的鮮葉中兒茶素與氨基酸比值也會(huì)保持在一定的比例范圍。鮮葉如萎凋不足,葉質(zhì)硬脆,揉捻時(shí)芽葉易斷碎,茶汁易流失,成品茶香氣低,有苦澀味;鮮葉如萎凋過度,葉質(zhì)干硬,成品茶色澤灰枯,葉底花雜[1],鮮葉只有萎凋適度才能加工出色、香、味俱佳的紅茶。目前,對(duì)于紅茶萎凋程度的判斷主要依賴人工味覺、嗅覺等經(jīng)驗(yàn)判別,其結(jié)果容易受到判別者心情、身體狀況以及性別等諸多因素的影響,缺乏量化的指標(biāo),難以進(jìn)行在線快速判別。

高光譜圖像技術(shù)融合了光譜信息和圖像信息,既能利用光譜信息分析樣品的內(nèi)部品質(zhì)信息,也能基于圖像信息表征樣品的外部品質(zhì)特征[2]。近年來,高光譜圖像技術(shù)在農(nóng)業(yè)生產(chǎn)方面得到了廣泛應(yīng)用[3-7]。高光譜圖像技術(shù)在茶葉上應(yīng)用主要集中在茶葉分類、品質(zhì)檢測(cè)及茶園管理等方面。章海亮等融合光譜主成分信息和圖像信息實(shí)現(xiàn)了6種名優(yōu)綠茶的有效鑒別[8]。陳孝敬等也通過分析茶葉顏色特征,實(shí)現(xiàn)了茶葉分類[9]。蔣帆等利用高光譜技術(shù)結(jié)合支持向量機(jī)分類算法對(duì)機(jī)炒龍井茶等級(jí)進(jìn)行了分類識(shí)別[10]。趙杰文等利用高光譜技術(shù)對(duì)茶樹葉片中葉綠素含量及分布規(guī)律進(jìn)行了研究,并建立了茶樹缺氮診斷模型[11]。王曉慶等分析了炭疽病脅迫下的茶樹葉片高光譜特征,為茶園病害的預(yù)防提供了可能[12]。謝傳奇等利用高光譜圖像技術(shù)實(shí)現(xiàn)了茶葉在干燥過程中葉色和含水量的實(shí)時(shí)檢測(cè)[13-14]。萎凋是紅茶加工的首要工序,對(duì)萎凋程度的判斷直接影響后續(xù)加工工序的進(jìn)行,將高光譜圖像技術(shù)應(yīng)用到判別紅茶萎凋程度的研究,未見相關(guān)報(bào)道。

本研究采集不同萎凋程度樣本的高光譜圖像,利用主成分分析法優(yōu)選特征波長(zhǎng)和對(duì)應(yīng)的光譜特征值,并提取特征波長(zhǎng)下的圖像紋理值,融合光譜和紋理特征值,結(jié)合化學(xué)計(jì)量學(xué)方法建立紅茶萎凋程度的快速判別模型,并嘗試分析萎凋程度和兒茶素與氨基酸比值的關(guān)系,實(shí)現(xiàn)對(duì)紅茶萎凋程度量化判斷。

1 材料與方法

1.1 試驗(yàn)材料

試驗(yàn)材料取自安徽祁門金東茶廠2015年4月8日和2016年4月19日生產(chǎn)樣。萎凋方式為萎凋槽萎凋,萎凋熱風(fēng)溫度30~32 ℃,攤?cè)~厚度為10~15 cm。萎凋1 h后取若干樣品作為萎凋不足樣,萎凋適度取樣后,繼續(xù)萎凋至過度,萎凋適度的判別由兩名經(jīng)驗(yàn)豐富的師傅和1名專家共同決定。達(dá)到萎凋適度時(shí),萎凋葉表面失去光澤,葉色暗綠,青草氣減退;葉形皺縮,葉質(zhì)柔軟,折梗不斷,緊握成團(tuán),松手可緩慢松散。最終獲得168個(gè)樣品,其中萎凋不足樣品55個(gè),萎凋適度樣品61個(gè),萎凋過度樣品52個(gè)。按照2∶1的比例將樣品隨機(jī)分成校正集(112個(gè))和預(yù)測(cè)集(56個(gè)),利用校正集的樣品建立判別模型,預(yù)測(cè)集的樣品用來測(cè)試模型的性能。

1.2 高光譜圖像采集和處理

高光譜圖像系統(tǒng)主要由光譜成像儀(Imspector V17E, Spectral Imaging Ltd., Oulu, Finland)、2個(gè)150 W的鹵素?zé)簦?900, Illumination Technologies Inc., New York, USA)、移動(dòng)平臺(tái)、暗箱、圖像采集和分析軟件(Spectral Image Software, Isuzu Optics Corp., Taiwan, China)等組成。

為了保證采集的圖像的質(zhì)量,首先要對(duì)相機(jī)的曝光時(shí)間以及移動(dòng)平臺(tái)的移動(dòng)速度等參數(shù)進(jìn)行設(shè)置。經(jīng)過反復(fù)調(diào)節(jié),最終曝光時(shí)間設(shè)置為2 ms,物鏡的高度設(shè)為31 cm。稱取(15±0.5) g的樣品均勻平鋪在規(guī)格為×h:9 cm×1 cm的培養(yǎng)皿中,置于移動(dòng)平臺(tái)上以9.5 mm/s的速度開始采集高光譜圖像。系統(tǒng)的光譜分辨率為5 nm,光譜范圍為908~1 735 nm,共508個(gè)波段。而前后波段受噪聲影響較大,因此在后續(xù)的數(shù)據(jù)處理過程中,選取947~1 696 nm波段范圍內(nèi),共448個(gè)波段的高光譜圖像進(jìn)行分析。

在對(duì)高光譜圖像處理前,先要按照式(1)對(duì)像素點(diǎn)進(jìn)行黑白校正。

R=(1)

式中R表示校正后的圖像像素點(diǎn),表示原始的圖像像素點(diǎn),表示黑板校正的圖像像素點(diǎn),表示白板校正的圖像像素點(diǎn)。

1.3 化學(xué)成分檢測(cè)

樣品進(jìn)行高光譜圖像采集后,立即微波固樣2 min,并在105 ℃的烘箱中烘至足干。茶葉樣經(jīng)磨碎后過篩至于密封袋中放在4 ℃的條件下備用。檢測(cè)的化學(xué)成分為兒茶素和氨基酸,兒茶素檢測(cè)采用ISO 14502-2方法,氨基酸的檢測(cè)采用Waters公司柱前衍生化法[15],共檢測(cè)18種氨基酸,并計(jì)算兒茶素總量和氨基酸的總量及兩者的比值,同時(shí)利用SPSS進(jìn)行相關(guān)的統(tǒng)計(jì)分析。

1.4 數(shù)據(jù)分析方法

1.4.1 特征波長(zhǎng)選擇

高光譜數(shù)據(jù)量龐大,且相鄰波段間相關(guān)性強(qiáng)易造成冗余的數(shù)據(jù),會(huì)影響后期數(shù)據(jù)的處理。因此,對(duì)高光譜數(shù)據(jù)進(jìn)行降維,去除冗余信息,優(yōu)選特征波長(zhǎng)是十分必要的。主成分分析法[16](principal component analysis,PCA)是一種常用的數(shù)據(jù)降維方法,主要是通過協(xié)方差最大的方向?qū)⒏呔S數(shù)據(jù)空間向低維數(shù)據(jù)空間投影,將原始數(shù)據(jù)轉(zhuǎn)化到新的坐標(biāo)系統(tǒng)中,得到的各主成分變量彼此間不相關(guān),且都是原始數(shù)據(jù)的線性組合,如公式(2)所示

PC=(2)

式中PC表示第個(gè)主成分圖像,表示原始圖像的波段數(shù),α表示第個(gè)波段處圖像的權(quán)重系數(shù),I表示第個(gè)波段處的原始圖像。本研究根據(jù)方差貢獻(xiàn)率提取主成分圖像,并通過比較主成分圖像下各波長(zhǎng)的權(quán)重系數(shù)的絕對(duì)值大小來優(yōu)選特征波長(zhǎng)。

1.4.2 紋理特征提取

基于灰度共生矩陣(grey level co-occurrence matrix, GLCM)來提取特征波長(zhǎng)圖像的紋理值?;叶裙采仃嘯17]是通過計(jì)算特定像素間距離和角度的函數(shù)。本研究中,距離設(shè)置為1,提取0°、45°、90°和135°4個(gè)角度的對(duì)比度、相關(guān)性、能量和同質(zhì)性4個(gè)紋理變量。利用連續(xù)投影算法(successive projections algorithm,SPA)對(duì)紋理特征進(jìn)行優(yōu)選,減少變量間的冗余信息。連續(xù)投影算法[18-19]是一種前向特征變量選擇算法,在優(yōu)選光譜變量時(shí)應(yīng)用廣泛[20-21],本研究用于紋理變量的篩選,以校正集樣本的紋理特征和類別賦值為輸入,并以預(yù)測(cè)集的紋理特征和類別作為驗(yàn)證,設(shè)置選擇的變量數(shù)范圍為1~20。

1.4.3 建模方法選擇

線性判別分析法[22](linear discriminat analysis,LDA)是模式識(shí)別中常用的線性分類方法。支持向量機(jī)[23](support vector machine,SVM)的主要思想是建立一個(gè)分類超平面為決策曲面,在分類問題上能提供較好的泛化性能。本研究的SVM實(shí)現(xiàn)采用的是LIBSVM工具箱[24],設(shè)置為RBF核函數(shù),通過交互驗(yàn)證尋找最佳的懲罰參數(shù)c和核函數(shù)參數(shù)g。極限學(xué)習(xí)機(jī)[25](extreme learning machine,ELM)是針對(duì)單隱含前饋神經(jīng)網(wǎng)絡(luò)(single hidden layer feedforward neural network,SLFN)的算法,只需要設(shè)置隱含層的神經(jīng)元個(gè)數(shù)便能得到最優(yōu)的唯一解。偏最小二乘法[26](partial least squares,PLS)是結(jié)合主成分分析和多元線性回歸的化學(xué)計(jì)量學(xué)方法,需要設(shè)置合理的PLS因子數(shù)來達(dá)到最佳的模型效果。不同方法建立的模型,判別結(jié)果可能也會(huì)存在差異。

1.5 數(shù)據(jù)處理

采用ENVI 4.7(ITT Visual Information Solutions, Boulder, USA),Matlab 2014a(The Mathworks Inc., Massachusetts, USA)和IBM SPSS(Version 20.0, Inc., Chicago, IL, USA)軟件對(duì)數(shù)據(jù)進(jìn)行處理。

2 結(jié)果與分析

2.1 不同萎凋程度樣品兒茶素與氨基酸比值分析

不同萎凋程度樣品兒茶素與氨基酸比值的統(tǒng)計(jì)結(jié)果如表1所示。從表1可以看出,適度萎凋鮮葉的兒茶素與氨基酸的平均比值在3.64左右,不同萎凋程度樣品的兒茶素與氨基酸比值之間重疊較少,因而可以利用兒茶素與氨基酸比值來區(qū)分萎凋程度。

茶鮮葉在萎凋過程中隨著水分的減少,細(xì)胞液濃縮,鮮葉內(nèi)源酶濃度增加,酶活性增強(qiáng)。一方面兒茶素類部分降解,同時(shí)蛋白水解酶活性增加,部分水溶性蛋白分解,氨基酸含量增加,因而兒茶素總量呈下降趨勢(shì),氨基酸含量呈現(xiàn)增加的趨勢(shì)[27]。本研究計(jì)算出兒茶素與氨基酸比值,萎凋不足樣品的兒茶素與氨基酸比值集中分布在5~4之間,萎凋適度樣品的兒茶素與氨基酸比值集中分布在4~3之間,萎凋過度樣品的兒茶素與氨基酸比值集中分布在3~2之間,這與劉少群等[28]分析的單芽型紅茶在萎凋過程中兒茶素與氨基酸的變化結(jié)果一致。

表1 兒茶素與氨基酸比值統(tǒng)計(jì)結(jié)果

2.2 光譜特征波長(zhǎng)的篩選

選擇校正后的圖像中間100×100像素范圍為感興趣區(qū)域(region of interest, ROI),提取ROI區(qū)域的所有像素的光譜平均值作為該樣本的光譜值。計(jì)算3種萎凋程度樣本的平均值,樣品的原始光譜如圖1所示,不同萎凋程度樣本的光譜曲線趨勢(shì)相似,在1 182和1 449 nm處出現(xiàn)明顯的吸收峰。為了消除光譜中的噪聲信息,采用標(biāo)準(zhǔn)正態(tài)變量法(standard normal variate, SNV)對(duì)原始光譜進(jìn)行預(yù)處理[29]。

對(duì)不同萎凋程度樣本的圖像進(jìn)行主成分分析,不同主成分?jǐn)?shù)對(duì)應(yīng)的方差貢獻(xiàn)率結(jié)果如表2所示。

表2 前5個(gè)主成分的方差貢獻(xiàn)率

由表2可知,第一主成分(principal component1, PC1)的方差貢獻(xiàn)率為96.66%,PC2的方差貢獻(xiàn)率為2.93%,前2個(gè)主成分的累積方差貢獻(xiàn)率達(dá)99.59%,幾乎可以代表全部信息。由于前2個(gè)主成分能解釋原始數(shù)據(jù)的絕大部分信息,因而可以利用前2個(gè)主成分來尋找特征波長(zhǎng)。如圖2所示,根據(jù)前2個(gè)主成分圖像下各波長(zhǎng)的權(quán)重系數(shù)的絕對(duì)值的大小優(yōu)選5個(gè)特征波長(zhǎng),分別是1 040、1 182、1 249、1 449、1 655 nm。

2.3 圖像紋理特征值分析

基于灰度共生矩陣從5個(gè)特征波長(zhǎng)圖像分別提取4個(gè)角度的對(duì)比度、相關(guān)性、能量和同質(zhì)性,共提取了80個(gè)紋理特征值。為進(jìn)一步分析紋理特征值中哪些特征值與萎凋程度關(guān)系更為密切,本研究采用SPA算法來優(yōu)選紋理特征值,當(dāng)變量個(gè)數(shù)為14時(shí),均方根誤差(root mean square error, RMSE)最小,為0.421 09,所以最終優(yōu)選出14個(gè)紋理特征值,分別為:1 040 nm處圖像下45°的相關(guān)性和135°的對(duì)比度,1 182 nm處圖像下0°的對(duì)比度,1 249 nm處圖像下45°和135°的對(duì)比度、90°的相關(guān)性和能量,1 449 nm處圖像的0、45°、90°和135°的對(duì)比度、45°的相關(guān)性,1 655 nm處圖像下的0°的能量和45°的對(duì)比度。由優(yōu)選得到的紋理特征值可知,紋理特征值主要集中在灰度共生矩陣的對(duì)比度和相關(guān)性上。

2.4 萎凋程度定性判別模型建立

光譜特征能表征萎凋葉的內(nèi)部品質(zhì),紋理特征能表現(xiàn)萎凋葉的外部特點(diǎn),為了更好的表示萎凋程度的變化,本研究將優(yōu)選的5個(gè)光譜特征值和14個(gè)紋理特征值在特征層[30]進(jìn)行融合。分別將光譜特征值、紋理特征值以及光譜和紋理特征值融合的數(shù)據(jù)作為L(zhǎng)DA、SVM和ELM模型的輸入值,建立萎凋程度判別模型,結(jié)果如表3所示。從表3可以看出,單獨(dú)的光譜特征值為模型的輸入值時(shí),LDA和SVM模型預(yù)測(cè)集的判別率都為91.07%,略高于ELM模型預(yù)測(cè)集的判別率。單獨(dú)的紋理特征值為模型的輸入值時(shí),LDA模型預(yù)測(cè)集的判別率為89.29%,高于ELM和SVM模型預(yù)測(cè)集的判別率。比較單獨(dú)的光譜特征值和紋理特征值建立的模型,可知基于光譜特征值建立的模型優(yōu)于基于紋理特征值建立的模型?;诠庾V和紋理特征值融合的數(shù)據(jù)的LDA模型,預(yù)測(cè)集的判別率最高,如表4所示,總體判別率達(dá)到94.64%,有1個(gè)萎凋不足的樣本被誤判為萎凋適度、1個(gè)萎凋適度的樣本被誤判為萎凋不足、1個(gè)萎凋過度的樣本被誤判為萎凋適度,可見誤判發(fā)生在相鄰程度的樣本間。通過比較可知,綜合光譜和紋理特征值的融合數(shù)據(jù)為模型輸入時(shí),模型的效果優(yōu)于基于單一特征值建立的模型。這可能是融合數(shù)據(jù)綜合萎凋葉內(nèi)外部的特征,能更加全面的反映出萎凋葉的變化。

2.5 萎凋程度定量預(yù)測(cè)模型建立

鮮葉在萎凋的過程中,隨著含水量的減少,兒茶素逐步降解,同時(shí)氨基酸會(huì)少量增加。在萎凋過程中,兒茶素與氨基酸比值一直處于變化之中,萎凋適度鮮葉的比值會(huì)相對(duì)穩(wěn)定地分布在一定范圍內(nèi),所以也可以通過定量檢測(cè)兒茶素與氨基酸比值來判別萎凋程度。為了更加準(zhǔn)確地判別紅茶的萎凋程度,對(duì)萎凋后茶樣中兒茶素與氨基酸比值進(jìn)行了定量預(yù)測(cè)。采用偏最小二乘法基于光譜和紋理特征值融合的數(shù)據(jù)建立兒茶素與氨基酸比值的預(yù)測(cè)模型,以相關(guān)系數(shù)(correlation coefficient,),驗(yàn)證均方根誤差(root mean square error for validation, RMSECV),預(yù)測(cè)均方根誤差(root mean square error of prediction, RMSEP)為模型的評(píng)價(jià)指標(biāo),模型的評(píng)價(jià)結(jié)果如圖3所示。預(yù)測(cè)集的p=0.8765,RMSEP=0.434,這表示優(yōu)選出的融合數(shù)據(jù)用來預(yù)測(cè)兒茶素與氨基酸比值是可行的。我國(guó)茶區(qū)分布較廣,不同茶區(qū)、不同茶樹品種鮮葉之間內(nèi)含成分存在著差異,因而,預(yù)測(cè)模型的建立需要針對(duì)某一茶類,同時(shí)模型的性能還需要不斷地優(yōu)化。

表3 不同特征值模型對(duì)萎凋程度的識(shí)別結(jié)果

注:LDA、SVM和ELM分別表示線性判別分析法、支持向量機(jī)和極限學(xué)習(xí)機(jī),下同。

Note: LDA, SVM and ELM represent linear discriminat analysis, support vector machine and extreme learning machine respectively, the same as below.

表4 光譜和紋理信息融合的LDA模型預(yù)測(cè)集的判別率

3 結(jié) 論

本研究基于光譜信息和圖像信息融合技術(shù)結(jié)合模式識(shí)別,判別紅茶的萎凋程度。通過主成分分析法優(yōu)選5個(gè)特征波長(zhǎng)及對(duì)應(yīng)光譜特征值,利用灰度共生矩陣提取5個(gè)特征波長(zhǎng)下的圖像紋理特征值,并通過連續(xù)投影算法優(yōu)選出14個(gè)紋理特征值,融合光譜和紋理特征值結(jié)合LDA模型算法,建立紅茶萎凋程度的判別快速模型,模型的判別率分別達(dá)到94.64%。模型的輸入值僅為19個(gè)變量,大大減少了建模所需的時(shí)間。利用光譜和紋理特征值融合的數(shù)據(jù)建立兒茶素與氨基酸的比值的定量預(yù)測(cè)模型,相關(guān)系數(shù)為0.8765,均方根誤差為0.434,預(yù)測(cè)結(jié)果較好。

[1] 安徽農(nóng)業(yè)大學(xué). 制茶學(xué)(第3版)[M]. 北京:中國(guó)農(nóng)業(yè)出版社,2016.

[2] Gowen A A, O’Donnell C P, Cullen P J, et al. Hyperspectral imaging-an emerging process analytical tool for food quality and safety control[J]. Trends in Food Science and Technology, 2007, 18: 590-598.

[3] 張紅濤,毛罕平,韓綠化. 近紅外高光譜成像技術(shù)檢測(cè)糧倉(cāng)米象活蟲[J]. 農(nóng)業(yè)工程學(xué)報(bào),2012,28(8):263-268.

Zhang Hongtao, Mao Hanping, Han Lühua. Detection of live(L.) in stored wheat by near- infrared hyperspectral imaging[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2012, 28(8): 263-268. (in Chinese with English abstract)

[4] 郭澎濤,蘇藝,茶正早,等. 基于BP神經(jīng)網(wǎng)絡(luò)的橡膠苗葉片磷含量高光譜預(yù)測(cè)[J]. 農(nóng)業(yè)工程學(xué)報(bào),2016,32(增刊1):177-183.

Guo Pengtao, Su Yi, Cha Zhengzao, et al. Prediction of leaf phosphorus contents for rubber seedlings based on hyperspectral sensitive bands and back propagation artificial neural network[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(Supp.1): 177-183. (in Chinese with English abstract)

[5] 陳志強(qiáng),王磊,白由路,等. 玉米葉片的光譜響應(yīng)及其氮素含量預(yù)測(cè)研究[J]. 光譜學(xué)與光譜分析,2013,33(4):1066-1070.

Chen Zhiqiang, Wang Lei, Bai Youlu, et al. Spectral response of maize leaves and prediction of their nitrogen content[J]. Spectroscopy and Spectral Analysis, 2013, 33(4): 1066-1070. (in Chinese with English abstract)

[6] 李金夢(mèng),葉旭君,王巧男,等. 高光譜成像技術(shù)的柑橘植株葉片含氮量預(yù)測(cè)模型[J]. 光譜學(xué)與光譜分析,2014,34(1):212-216.

Li Jinmeng, Ye Xujun, Wang Qiaonan, et al. Development of prediction models for determining N content in citrus leaves based on hyperspectral imaging technology[J]. Spectroscopy and Spectral Analysis, 2014, 34(1): 212-216. (in Chinese with English abstract)

[7] 張瑤,鄭立華,李民贊,等. 基于光譜學(xué)原理與小波包分解技術(shù)預(yù)測(cè)蘋果樹葉片氮素含量[J]. 農(nóng)業(yè)工程學(xué)報(bào),2013,29(增刊1):101-108.

Zhang Yao, Zheng Lihua, Li Minzan, et al. Predicting apple leaf nitrogen content based on hyperspectral and wavelet packet analysis[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2013, 29(Supp.1): 101-108. (in Chinese with English abstract)

[8] 章海亮,李曉麗,朱逢樂,等. 應(yīng)用高光譜成像技術(shù)鑒別綠茶品牌研究[J].光譜學(xué)與光譜分析,2014,34(5):1373-1377.

Zhang Hailiang, Liu Xiaoli, Zhu Fengle, et al. Identification of green tea brand based on hyperspectral imaging[J]. Spectroscopy and Spectral Analysis, 2014, 34(5): 1373-1377. (in Chinese with English abstract)

[9] 陳孝敬,吳迪,何勇,等. 基于多光譜圖像顏色特征的茶葉分類研究[J]. 光譜學(xué)與光譜分析,2008,28(11):2527-2530.

Chen Xiaojing, Wu Di, He Yong et al. Study on discrimination of tea based on color of multispectral image[J]. Spectroscopy and Spectral Analysis, 2008, 28(11): 2527-2530. (in Chinese with English abstract)

[10] 蔣帆,喬欣,鄭華軍,等.基于高光譜分析技術(shù)的機(jī)炒龍井茶等級(jí)識(shí)別方法[J]. 農(nóng)業(yè)工程學(xué)報(bào),2011,27(7):343-348.

Jiang Fan, Qiao Xin, Zheng Huajun, et al. Grade discrimination of machine fried Longjing tea based on hyperspectral technology[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2011, 27(7): 343-348. (in Chinese with English abstract)

[11] 趙杰文,王開亮,歐陽琴,等. 高光譜技術(shù)分析茶樹葉片中葉綠素含量及分布[J]. 光譜學(xué)與光譜分析,2011(2):512-515.

Zhao Jiewen, Wang Kailiang, Ouyang Qin, et al. Measurement of chlorophyll content and distribution in tea plant’s leaf using hyperspectral imaging technique[J]. Spectroscopy and Spectral Analysis, 2011(2): 512-515. (in Chinese with English abstract)

[12] 王曉慶,冉烈,彭萍,等. 炭疽病脅迫下的茶樹葉片高光譜特征分析[J].植物保護(hù),2014,40(6):13-17.

Wang Xiaoqing, Ran Lie, Peng Ping, et al. Analysis of the hyperspectral characteristics of tea leaves under anthracnose disease stress[J]. Plant Protection, 2014, 40(6): 13-17. (in Chinese with English abstract)

[13] Xie Chuanqi, Li Xiaoli, Shao Yongni, et al. Color measurement of tea leaves at different drying periods using hyperspectral imaging technique[J]. PLOS ONE, 2014, 9(12): e113422.

[14] Xie Chuanqi, Li Xiaoli , He yong, et al. Application of time series hyperspectral imaging (TS-HSI) for determining water content within tea leaves during dying[J]. Transactions of the ASABE, 2013, 56(6): 1431-1440.

[15] 寧井銘,顏玲,張正竹,等. 祁門紅茶加工中氨基酸和兒茶素快速檢測(cè)模型建立[J]. 光譜學(xué)與光譜分析,2015(12):3422-3426.

Ning Jingming, Yan Ling, Zhang Zhengzhu, et al. Rapid and dynamic determination models of amino acids and catechins concentrations during the processing procedures of keemun balck tea[J]. Spectroscopy and Spectral Analysis, 2015(12): 3422-3426. (in Chinese with English abstract)

[16] 吳龍國(guó),王松磊,康寧波,等. 基于高光譜成像技術(shù)的靈武長(zhǎng)棗缺陷識(shí)別[J]. 農(nóng)業(yè)工程學(xué)報(bào),2015,31(20):281-286.

Wu Longguo, Wang Songlei, Kang Ningbo, et al. Non-destructive detection of defects inMill cv.based on near-infrared hyperspectral imaging[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(20): 281-286. (in Chinese with English abstract)

[17] Haralick R M, Shanmugam K, Dinstein I. Textural features for image classification[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1973, 3(6): 610-621.

[18] Soares S F C, Gomes A A, Galvao A R, et al. The successive projections algorithm[J]. TRAC-Trends in Analytical Chemisty, 2013, 42(42): 84-98.

[19] 金瑞,李小昱,顏伊蕓,等. 基于高光譜圖像和光譜信息融合的馬鈴薯多指標(biāo)檢測(cè)方法[J]. 農(nóng)業(yè)工程學(xué)報(bào),2015,31(16):258-263.

Jin Rui, Li Xiaoyu, Yan Yiyun, et al. Detection method of multi-target recognition of potato based on fusion of hyperspectral imaging and spectral information[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(16): 258-263. (in Chinese with English abstract)

[20] Shao Y N, Li Y, Jiang, L J, et al. Identification of pesticide varieties by detecting characteristics ofusing Visible/Near infrared hyperspectral imaging and Raman microspectroscopy technology[J]. Water Research, 2016, 104: 432-440.

[21] 文韜,洪添勝,李立君,等. 基于高光譜技術(shù)的霉變稻谷脂肪酸含量無損檢測(cè)[J]. 農(nóng)業(yè)工程學(xué)報(bào),2015,31(18):233-239.

Wen Tao, Hong Tiansheng, Li Lijun, et al. Non-destructive detection of fatty acid content in mould paddy based on high-spectral technology[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(18): 233-239. (in Chinese with English abstract)

[22] Wei X, Liu F, Qiu Z J, et al. Ripeness classification of astringent persimmon using hyperspectral imaging technique[J]. Food Bioprocess Technology, 2014(7): 1371-1380.

[23] Wang L, Sun D W, Pu H B, et al. Application of hyperspectral imaging to discriminate the variety of maize seeds[J]. Food Analyticla Methods, 2016(9): 225-234.

[24] Chang C C, Lin C J. LIBSVM: a library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3): 389-396.

[25] Huang G, Huang G B, Song S J, et al. Trends in extreme learning machines: A review[J]. Neural Networks, 2015, 61: 32-48.

[26] 劉燕德,熊松盛,吳至境,等. 贛南臍橙園土壤全磷和全鉀近紅外光譜檢測(cè)[J]. 農(nóng)業(yè)工程學(xué)報(bào),2013,29(18):156-162.

Liu Yande, Xiong Songsheng, Wu Zhijing, et al. Detection of total potassium and total phosphorus in soil in GAN NAN navel orange orchard using near infrared spectroscopy[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2013, 29(18): 156-162. (in Chinese with English abstract)

[27] 黃藩,董春旺,朱宏凱,等. 紅茶萎凋中鮮葉理化變化及工藝技術(shù)研究進(jìn)展[J]. 中國(guó)農(nóng)學(xué)通報(bào),2014,30(27):275-281.

Huang Fan, Dong Chunwang, Zhu hongkai, et al. A review on the black tea withering technology and the changes of physics and chemistry of leaves[J]. Chinese Agricultural Science Bulletin, 2014, 30(27): 275-281. (in Chinese with English abstract)

[28] 劉少群,孫彬妹,段瑩,等. 不同萎凋程度處理對(duì)單芽紅茶品質(zhì)的影響[J]. 茶葉通訊,2015,42(4):30-34.

Liu Shaoqun, Sun Binmei, DuanYing, et al. Effect of withering degree on the quality of single bud black tea[J]. Journal of Tea Communication, 2015, 42(4): 30-34. (in Chinese with English abstract)

[29] Huang L, Zhao J W, Chen Q S, et al. Rapid detection of total viable count (TVC) in pork meat by hyperspectral imaging[J]. Food Research International, 2013, 54(1): 821-828.

[30] Xiong Z J, Sun D W, Pu H B, et al. Combination of spectra and texture data of hyperspectral imaging for differentiating between free-range and broiler chicken meats[J]. LWT-Food Science and Technology, 2015, 60(2): 649-655.

Discriminant of withering quality of Keemun black tea based on information fusion of image and spectrum

Ning jingming1, Sun Jingjing1, Zhu Xiaoyuan1, Li Shuhuan1, Zhang Zhengzhu1, Huang Caiwang2

(1.230036;2.245600,)

Withering is the first procedure and the key step in processing of black tea. It is crucial for the quality of black tea product. Usually, the judgment of the withering degree relies on the processor’s judgment, rather than a quantitative analysis by fast evaluation method. In order to develop the digitized discrimination on withering degrees, different degrees of withering samples were collected in our research. In this study, 168 samples provided by Jindong tea factory in Qimen County were investigated. All of the samples belonged to different withering degrees (55 samples of mild withering, 61 samples of moderate withering and 52 samples of excessive withering). The samples were randomly divided into two subsets at the ratio of 2:1. 112 samples were chosen as the calibration set and the remaining 56 samples were prediction set. The calibration set was used to develop the model, while the prediction set was applied to test the robustness of the model. The withering degree was nondestructively evaluated by hyperspectral imaging technology at the range of 908-1735 nm. It was suggested that the ratio of catechins/amino acids was correspondingly decreased with the development of withering degrees. Furthermore, the contents of catechins and amino acids of these samples were detected by high-performance liquid chromatography (HPLC). The characteristic spectra were extracted from the region of interest (ROI), and standard normal variate (SNV) method was preprocessed to reduce background noise. All of the hyperspectral images of tea samples with different withering degrees were analyzed by principal component analysis (PCA). The first two principal component (PC) images were selected because PC1 and PC2 contributed to 99.59% variance of the total. Therefore, the first two PC images were used for selecting dominate wavelengths. And five dominant wavelengths (1 040, 1 182, 1 249, 1 449 and 1 655 nm) were selected as spectral features. Textual features were collected by Grey level co-occurrence matrix (GLCM) from five dominant wavelengths of images. Fourteen dominant textual features were selected by successive projections algorithm (SPA). Subsequently, linear discriminant analysis (LDA), support vector machine (SVM) and extreme learning machine (ELM) classification models were developed based on spectral features, textural features and data fusion, respectively. Compared with the results of the models built with spectral features or textural features, the LDA, SVM and ELM models based on data fusion showed higher correct discrimination rate in prediction set. The correct discrimination rate of LDA, SVM and ELM based on data fusion were 94.64%, 91.07% and 92.86%, respectively. The results indicated that hyperspectral imaging combined with LDA was a potent tool in the discrimination of withering degrees. At the same time, catechins/amino acids ratio was also applied in the discrimination of withering degrees. The study showed that correlate coefficient of prediction set by catechins/amino acids ratio was 0.8765, and root mean square error of prediction was 0.434. The results in this study provide a new method with fast and scientific of digitized discrimination for withering degree during black tea processing.

data fusion; discriminant analysis; image analysis; partial least squares approximations; black tea; withering; ratio of catechins to amino acids

10.11975/j.issn.1002-6819.2016.24.041

TS272.7; S123

A

1002-6819(2016)-24-0303-06

2016-09-30

2016-11-17

國(guó)家重點(diǎn)研發(fā)計(jì)劃(2016YFD0200900);國(guó)家現(xiàn)代農(nóng)業(yè)(茶葉)產(chǎn)業(yè)體系(CARS-23)

寧井銘,男,漢族,安徽泗縣人,副教授,博士,主要從事茶葉加工、茶葉品質(zhì)分析及紅外光譜技術(shù)在茶葉上應(yīng)用研究。合肥 茶樹生物學(xué)與資源利用國(guó)家重點(diǎn)實(shí)驗(yàn)室,安徽農(nóng)業(yè)大學(xué),230036。Email:ningjm@ahau.edu.cn

猜你喜歡
兒茶素紅茶特征值
《幸福的紅茶時(shí)光》
食品界(2022年11期)2022-12-12 08:04:44
一類帶強(qiáng)制位勢(shì)的p-Laplace特征值問題
超高效液相色譜法測(cè)定茶葉中的兒茶素
單圈圖關(guān)聯(lián)矩陣的特征值
蜜香紅茶
貴茶(2019年3期)2019-12-02 01:47:22
紅茶與綠茶有什么區(qū)別?
基于商奇異值分解的一類二次特征值反問題
關(guān)于兩個(gè)M-矩陣Hadamard積的特征值的新估計(jì)
全甲基化沒食子兒茶素沒食子酸酯的制備
兒茶素酶促制備茶黃素的研究進(jìn)展
茶葉通訊(2014年2期)2014-02-27 07:55:38
广德县| 宝鸡市| 嘉荫县| 衢州市| 佛教| 泽库县| 东阿县| 高邮市| 绿春县| 巴南区| 和田县| 浮梁县| 文化| 湘乡市| 专栏| 都昌县| 淅川县| 新邵县| 瑞丽市| 师宗县| 大埔区| 临猗县| 连南| 麟游县| 东阿县| 根河市| 秦安县| 和硕县| 德兴市| 博客| 堆龙德庆县| 丽江市| 多伦县| 张家界市| 洛浦县| 临海市| 沙洋县| 岐山县| 元江| 镇原县| 阿克苏市|