張萬(wàn)紅，劉文兆

覆膜玉米冠層圖像分割方法

張萬(wàn)紅，劉文兆*1

（西北農(nóng)林科技大學(xué)水土保持研究所，陜西楊凌712100）

在弱光條件下，采用色調(diào)（H）和飽和度（S）顏色分量的K均值聚類(lèi)分析結(jié)合相應(yīng)色差運(yùn)算方法，對(duì)覆膜玉米冠層圖像進(jìn)行分割，并將分割所得影像的二值圖分別與超綠、超紅和超綠－超紅算法分割結(jié)果進(jìn)行比較。結(jié)果表明，該方法更能精確反映玉米的冠層形狀。將該方法得到的玉米冠層覆蓋度計(jì)算結(jié)果與Samplepoint軟件分析結(jié)果進(jìn)行比較發(fā)現(xiàn)，前者均方根誤差取值較小，僅為0.004 2,分割誤差率低至3.37%，分割圖像準(zhǔn)確率高。綜合分析表明,在弱光背景下，基于H和S顏色分量的K均值聚類(lèi)分析結(jié)合色差運(yùn)算的分割方法對(duì)覆膜玉米冠層的分割結(jié)果準(zhǔn)確可靠。

色調(diào)；飽和度；K均值聚類(lèi)；圖像分割；玉米冠層

植被的冠層覆蓋度是指植被在地面的垂直投影面積占統(tǒng)計(jì)區(qū)總面積的百分比[1]，它既能反映植物生長(zhǎng)期內(nèi)的動(dòng)態(tài)變化，又能間接說(shuō)明植物的蒸騰作用和光合作用[2]。因此，在農(nóng)田條件下準(zhǔn)確地估測(cè)植被的冠層覆蓋度對(duì)監(jiān)測(cè)作物生長(zhǎng)狀態(tài)和預(yù)測(cè)作物產(chǎn)量具有重要意義[3-4]。

傳統(tǒng)估測(cè)作物冠層覆蓋度的方法有目測(cè)法、尺測(cè)法、逐點(diǎn)目視判斷并計(jì)數(shù)垂直成像照片法[2，5-6]。雖然這些方法很簡(jiǎn)單，但目測(cè)法主觀隨意性很強(qiáng)，不同觀測(cè)者可能結(jié)果不同，甚至相差很大；尺測(cè)法受天氣影響大，估測(cè)結(jié)果帶方向性，且勞動(dòng)強(qiáng)度大；逐點(diǎn)目視判斷并計(jì)數(shù)垂直成像照片法估測(cè)精確，但勞動(dòng)強(qiáng)度也大，耗費(fèi)時(shí)間長(zhǎng)[7]。

目前，利用數(shù)字相機(jī)拍攝作物冠層照片，在計(jì)算機(jī)中將影像分成作物和非作物（土壤、作物殘留物等），并利用二值圖像計(jì)算作物覆蓋度的方法操作簡(jiǎn)單，結(jié)果準(zhǔn)確率高，是一種適宜的方法[5]。對(duì)于背景簡(jiǎn)單的田間作物圖像，例如大田玉米圖像中背景僅包含土壤和少許作物殘留，通常采取單閾值的方法即可快速實(shí)現(xiàn)對(duì)目標(biāo)的識(shí)別與分割，但對(duì)于多背景的影像，單閾值分割法的準(zhǔn)確率低，往往會(huì)產(chǎn)生過(guò)度分割。為了實(shí)現(xiàn)對(duì)多背景目標(biāo)影像的準(zhǔn)確分割，通常將RGB顏色空間轉(zhuǎn)化為HIS[7-11]、HSV[12-13]和Lab[14-16]等顏色空間并結(jié)合最大類(lèi)間方差法（Otsu）[14]、K均值聚類(lèi)[15,17]和模糊C均值聚類(lèi)（fuzzy C-mean clustering,FCM）[18]等算法對(duì)圖像進(jìn)行分割。這些方法雖然能準(zhǔn)確將目標(biāo)影像分割，但目前仍沒(méi)有統(tǒng)一的算法來(lái)實(shí)現(xiàn)對(duì)不同環(huán)境條件下所有特定作物圖像的分割[19]。

地膜覆蓋可以改善農(nóng)田土壤的水熱狀況，提高養(yǎng)分有效性和水分利用效率，該種植方式已在玉米田得到廣泛推廣和應(yīng)用[20-21]。但目前將玉米植株從地膜、土壤等背景中分離并獲取玉米冠層覆蓋度的方法鮮有相關(guān)文獻(xiàn)報(bào)道?；诖?，本試驗(yàn)擬采用圖像處理方法將玉米植株從地膜、土壤等背景中分離，并最終實(shí)現(xiàn)對(duì)玉米冠層覆蓋度的準(zhǔn)確計(jì)算。此方法首先將覆膜玉米影像從RGB顏色空間轉(zhuǎn)換到HIS顏色空間，然后分別提取H和S顏色分量，通過(guò)對(duì)H和S顏色分量進(jìn)行K均值聚類(lèi)分析[22]，選取合適的2類(lèi)聚類(lèi)圖進(jìn)行相應(yīng)色差運(yùn)算，以分割覆膜玉米冠層圖像。將圖像分割結(jié)果分別與超綠（excess green,ExG）[23]、超紅（excess red,ExR）[24]以及超綠減超紅（ExG－ExR）[25]算法的分割結(jié)果進(jìn)行比較，在比較的基礎(chǔ)上選用合理的分割圖像進(jìn)行玉米冠層覆蓋度計(jì)算。

1 材料與方法

1.1 大田玉米圖像采集

試驗(yàn)于2016年6月4日—6日（玉米苗期）在中國(guó)科學(xué)院長(zhǎng)武黃土高原農(nóng)業(yè)生態(tài)試驗(yàn)站覆膜玉米試驗(yàn)田進(jìn)行。首先用1 m2的樣方框?qū)⒂衩字仓昕蚨?，然后使用華為榮耀7手機(jī)，在下午和早晨太陽(yáng)光較弱的時(shí)間段，采用自然曝光模式，垂直于每個(gè)框定樣方在地面上2 m處進(jìn)行拍照，收集光照均勻、少有陰影的圖片備用。在進(jìn)行圖像分割處理前，為方便圖像處理，在不影響圖像中目標(biāo)與背景形狀及顏色的前提下，將圖像統(tǒng)一變換為1 358×1 314像素，以JPEG格式導(dǎo)入計(jì)算機(jī)，如圖1所示。

圖1 覆膜玉米影像Fig.1 Original image of plastic-film corn

1.2 玉米植株圖像分割

覆膜玉米冠層圖像分割流程如圖2所示。

1）獲取RGB顏色空間下的R、G、B顏色分量，分別計(jì)算影像的超綠、超紅以及超綠－超紅算法結(jié)果；超綠、超紅以及超綠－超紅算法結(jié)合Otsu閾值分割算法[26]對(duì)圖像進(jìn)行分割。超綠及超紅算法如公式(1)～(2)[23-24]所示。

式中R、G、B分別代表紅、綠、藍(lán)顏色分量。

2）將RGB顏色空間轉(zhuǎn)換為HIS顏色空間[見(jiàn)公式(3)～(6)[10]],并提取H及S顏色分量；使用K均值聚類(lèi)算法分別對(duì)H及S顏色分量進(jìn)行聚類(lèi)分析（聚類(lèi)數(shù)為3），獲取2類(lèi)顏色分量的二值圖像，對(duì)二值圖進(jìn)行去噪和形態(tài)學(xué)開(kāi)運(yùn)算，斷開(kāi)細(xì)小黏連，去除毛刺使圖像更為平滑；對(duì)去噪及形態(tài)學(xué)運(yùn)算后的H和S顏色分量二值圖進(jìn)行相應(yīng)的數(shù)學(xué)運(yùn)算，獲取目標(biāo)圖像。

圖2 玉米冠層圖像分割方法流程圖Fig.2 Flow chart for image segmentation method of plastic-film corn canopy

1.3 數(shù)據(jù)處理

采用Excel 2013進(jìn)行數(shù)據(jù)處理及運(yùn)算。

2 結(jié)果與分析

2.1 超綠、超紅以及超綠－超紅算法結(jié)合Otsu閾值分割法

圖3為ExG算法圖像分割結(jié)果。該分割方法能準(zhǔn)確識(shí)別土壤、塑料膜、作物殘留等背景目標(biāo)，識(shí)別后的背景目標(biāo)在分割后的圖像中被標(biāo)識(shí)為黑色，但ExG算法不能準(zhǔn)確識(shí)別前景目標(biāo)（玉米）。在圖3中，部分玉米葉脈、葉尖以及下垂葉的葉緣區(qū)域呈現(xiàn)為黑色，表明ExG算法對(duì)前景目標(biāo)產(chǎn)生了過(guò)度分割。圖4為采用ExG算法獲取的圖像經(jīng)Otsu閾值分割法處理后的二值圖。從中可以看出，前景目標(biāo)被標(biāo)識(shí)為白色，背景目標(biāo)被標(biāo)識(shí)為黑色，前景目標(biāo)圖像中的零星黑色斑塊表明ExG算法結(jié)合Otsu閾值分割法同樣也產(chǎn)生了對(duì)圖像的過(guò)度分割現(xiàn)象。圖5為采用ExR算法獲取的分割圖像。在該圖中，除了葉片中有小部分區(qū)域與背景色基本一致外，大部分葉片的顏色與背景顏色呈現(xiàn)出明顯的差異，這種顏色差異更有利于對(duì)目標(biāo)和背景圖像進(jìn)行分割。圖6為采用Otsu閾值分割法對(duì)ExR算法獲取的圖像進(jìn)行自適應(yīng)閾值處理后的二值圖，圖中背景部分的土壤以及塑料膜在部分區(qū)域中呈現(xiàn)出與前景目標(biāo)一致的白色，說(shuō)明前景與背景的分割效果差。圖7與圖8分別為采用ExG－ExR算法以及ExG－ExR算法結(jié)合Otsu閾值分割法處理后的分割圖。這2種分割方法的分割結(jié)果分別與ExG以及ExG結(jié)合Otsu分割法的結(jié)果較為一致。

圖4 超綠算法結(jié)合自適應(yīng)閾值法分割的影像Fig.4 Image segmented by ExG and Otsu methods

圖5 超紅算法分割的影像Fig.5 Image segmented by ExR method

圖3 超綠算法分割的影像Fig.3 Image segmented by ExG method

圖6 超紅算法結(jié)合自適應(yīng)閾值法分割的影像Fig.6 Image segmented by ExR and Otsu methods

圖7 超綠減超紅算法分割的影像Fig.7 Image segmented by ExG－ExR method

圖8 超綠減超紅算法結(jié)合自適應(yīng)閾值法分割的影像Fig.8 Image segmented by ExG－ExR and Otsu methods

2.2 基于H和S顏色分量的K均值聚類(lèi)分割法

2.2.1 H和S顏色分量的K均值聚類(lèi)分析

獲取H和S顏色分量（圖9～10）后，對(duì)H和S顏色分量分別進(jìn)行K均值聚類(lèi)分析，然后選取適宜用于圖像分割的聚類(lèi)圖做后續(xù)分割處理?；贖顏色分量的K均值聚類(lèi)分析如圖11所示，基于S顏色分量的K均值聚類(lèi)分析如圖12所示。圖11準(zhǔn)確顯示了前景目標(biāo)的形狀，但在背景的塑料膜區(qū)域（圖1中間部分所示）中，由于塑料膜與土壤接觸的緊密程度以及膜厚度的不均一，導(dǎo)致凝結(jié)在塑料膜下的露珠區(qū)域呈現(xiàn)出斑塊狀與點(diǎn)狀交織在一起的白色，與土壤接觸緊密的區(qū)域部分呈現(xiàn)出接近干土的顏色，覆膜邊緣區(qū)域呈現(xiàn)出與裸露土壤接近的顏色。這種現(xiàn)象導(dǎo)致對(duì)圖像進(jìn)行聚類(lèi)分析后，前景目標(biāo)的部分葉邊緣出現(xiàn)過(guò)多噪點(diǎn)，小的葉縫隙間出現(xiàn)了黏連，增加了圖像分割的難度。但是，基于S顏色分量的K均值聚類(lèi)分析幾乎準(zhǔn)確呈現(xiàn)了覆膜背景區(qū)域（圖12中間部分的黑色區(qū)域），覆膜區(qū)域呈現(xiàn)干凈的黑色，很少有噪點(diǎn)產(chǎn)生，更重要的是，對(duì)比圖11中覆膜區(qū)域的玉米葉，圖12準(zhǔn)確呈現(xiàn)了覆膜區(qū)域復(fù)雜背景下的葉子形狀。

圖9 H顏色分量影像Fig.9 Image of hue

圖10 S顏色分量影像Fig.10 Image of saturation

2.2.2 運(yùn)用“Bwareaopen”程序及相應(yīng)數(shù)學(xué)運(yùn)算分割圖像

綜合以上分析，運(yùn)用Matlab軟件中的“Bwareaopen”命令對(duì)圖11中的噪點(diǎn)進(jìn)行清除。清除后的結(jié)果（圖13）顯示，圖中噪點(diǎn)清除很干凈，但前景目標(biāo)的部分葉子間隙有黏連且部分葉緣處有過(guò)多白色附著物，白色附著物與葉緣緊密結(jié)合在一起。為了消除葉緣附著物并恢復(fù)葉間隙，對(duì)圖12與圖13進(jìn)行減法運(yùn)算，再將運(yùn)算結(jié)果與圖13相減，通過(guò)減法運(yùn)算及去噪處理，部分葉子間的黏連以及葉緣處的附著物消失，顯示出了清晰的玉米冠層輪廓（圖14）。

圖11 H顏色分量的聚類(lèi)分析Fig.11 Clustering analysis for image of hue

圖12 S顏色分量的聚類(lèi)分析Fig.12 Clustering analysis for image of saturation

圖13 經(jīng)Bwareaopen軟件程序處理后的圖像Fig.13 Image treated by Bwareaopen program

圖14 最終分割的圖像Fig.14 Segmented image

2.3 實(shí)驗(yàn)結(jié)果及分析

為了驗(yàn)證算法的準(zhǔn)確性，運(yùn)用上述分割算法對(duì)采集到的20幅覆膜玉米圖像（每幅圖像代表的實(shí)際土地面積為1 m2）進(jìn)行分割，并根據(jù)計(jì)算公式(7)[27]和(8)分別計(jì)算玉米冠層圖像分割誤差率和均方根誤差（RMSE）。

式中：E為誤差率；C1為根據(jù)Samplepoint軟件（以人機(jī)交互的方式對(duì)土壤、植物、巖石等目標(biāo)物進(jìn)行判別）[28]測(cè)定的玉米冠層覆蓋度結(jié)果；C2為基于H和S顏色分量的K均值聚類(lèi)分析和色差運(yùn)算分割圖像后計(jì)算所得的玉米冠層覆蓋度；n為玉米冠層圖像數(shù)目。

計(jì)算結(jié)果顯示，RMSE取值較小，僅為0.004 2，誤差率低達(dá)3.37%：表明利用本文算法分割圖像后計(jì)算所得的玉米冠層覆蓋度與Samplepoint軟件測(cè)定結(jié)果非常接近，分割結(jié)果可靠。

3 討論與結(jié)論

準(zhǔn)確分割玉米冠層圖像對(duì)研究玉米生理生態(tài)具有重要意義。為了準(zhǔn)確分割覆膜條件下的玉米冠層圖像，本文提出了基于H和S顏色分量的K均值聚類(lèi)的算法，通過(guò)聚類(lèi)分析分別獲取基于H和S顏色分量的聚類(lèi)分析圖，根據(jù)2類(lèi)圖所反映的前景目標(biāo)及背景的差異，通過(guò)色差運(yùn)算的方法實(shí)現(xiàn)了對(duì)玉米冠層圖像的分割。為了證明這種方法對(duì)覆膜條件下玉米冠層圖像分割的有效性，分別選取ExG、ExR和ExG－ExR算法結(jié)合Otsu閾值分割法對(duì)覆膜玉米圖像進(jìn)行分割，將分割結(jié)果與基于H和S顏色分量聚類(lèi)分析的色差運(yùn)算分割結(jié)果進(jìn)行比較。結(jié)果顯示，基于H和S顏色分量K均值聚類(lèi)分析的色差運(yùn)算分割方法的分割結(jié)果優(yōu)于以上算法的分割結(jié)果。對(duì)基于H和S顏色分量聚類(lèi)分析的色差運(yùn)算分割結(jié)果進(jìn)行統(tǒng)計(jì)分析表明，基于H和S顏色分量K均值聚類(lèi)分析的色差運(yùn)算分割方法對(duì)覆膜玉米的分割誤差率低達(dá)3.37%，分割圖像準(zhǔn)確率高。綜上所述，基于H和S顏色分量的K均值聚類(lèi)分析結(jié)合相應(yīng)色差運(yùn)算的方法適宜對(duì)弱光條件下覆膜玉米影像進(jìn)行分割，且分割結(jié)果可靠。該方法對(duì)其他矮稈覆膜作物的冠層覆蓋度計(jì)算也具有一定的參考價(jià)值。

本研究是在弱光線(xiàn)條件下（如陰天、早晨或傍晚）進(jìn)行的，圖像的光照比較均勻，基于S顏色分量的聚類(lèi)分析補(bǔ)償了H顏色分量的聚類(lèi)分析對(duì)白色覆膜區(qū)域不能準(zhǔn)確識(shí)別的不足，同時(shí)，基于H顏色分量的聚類(lèi)分析則彌補(bǔ)了S顏色分量聚類(lèi)分析不能對(duì)綠色植株部分準(zhǔn)確識(shí)別的不足，兩者通過(guò)色差運(yùn)算實(shí)現(xiàn)了對(duì)覆膜玉米的準(zhǔn)確分割。但在強(qiáng)光線(xiàn)條件下，一些覆膜區(qū)域會(huì)產(chǎn)生陰影或反光，植株葉片受強(qiáng)光照射部分會(huì)產(chǎn)生反射光，土壤部分也存在明暗交替的光斑等現(xiàn)象，這些干擾項(xiàng)均會(huì)增加圖像分割的難度和不確定性。因此，本文提出的基于H和S顏色分量的K均值聚類(lèi)分析結(jié)合色差運(yùn)算的分割方法僅限于陰天、早晨或傍晚等弱光條件下進(jìn)行影像分割處理。

[1] GUEVARA-ESCOBAR A,TELLEZ J,GONZALEZ-SOSA E.Use of digital photography for analysis of canopy closure.Agroforestry Systems,2005,65(3):175-185.

[2] ADAMS J E,ARKIN G F.A light interception method for measuring row crop ground cover.Soil Science Society of America Journal,1977,41(4):789-792.

[3] 張學(xué)藝,郭建茂,韓穎娟,等.基于植被指數(shù)的寧夏灌區(qū)春小麥葉面積指數(shù)模型.中國(guó)農(nóng)業(yè)氣象,2011,32(2):279-282.ZHANG X Y,GUO J M,HAN Y J,et al.LAI model of spring wheat in Ningxia irrigated area based on MODIS-VI.Chinese Journal of Agrometeorology,2011,32(2):279-282.(in Chinese with English abstract)

[4] 瞿瑛,劉素紅,謝云.植被覆蓋度計(jì)算機(jī)模擬模型與參數(shù)敏感性分析.作物學(xué)報(bào),2008,34(11):1964-1969.QU Y,LIU S H,XIE Y.Computer simulation model of fractional vegetation cover and its parameters sensitivity.Acta Agronomica Sinica,2008,34(11):1964-1969.(in Chinese with English abstract)

[5] ARMBRUST D V.Rapid measurement of crop canopy cover.Agronomy Journal,1990,82(6):1170-1171.

[6] EWING R P,HORTON R.Quantitative color image analysis of agronomic images.Agronomy Journal,1999,91(1):148-153.

[7] 李存軍,王紀(jì)華,劉良云,等.基于數(shù)字照片特征的小麥覆蓋度自動(dòng)提取研究.浙江大學(xué)學(xué)報(bào)(農(nóng)業(yè)與生命科學(xué)版),2004,30(6):650-656.LI C J,WANG J H,LIU L Y,et al.Automated digital image analyses for estimating percent ground cover of winter wheat based on objectfeatures.JournalofZhejiang University(Agriculture and Life Sciences),2004,30(6):650-656.(in Chinese with English abstract)

[8] 黃芬,于琪,姚霞,等.小麥冠層圖像H分量的K均值聚類(lèi)分割.計(jì)算機(jī)工程與應(yīng)用,2014,50(3):129-134.HUANG F,YU Q,YAO X,et al.K-means clustering segmentation for H weight of wheat canopy image.Computer Engineering and Applications,2014,50(3):129-134.(in Chinese with English abstract)

[9] 李曉斌,王玉順,付麗紅.用K-means圖像法和主成分分析法監(jiān)測(cè)生菜生長(zhǎng)勢(shì)（英文）.農(nóng)業(yè)工程學(xué)報(bào),2016,32(12):179-186.LI X B,WANG Y S,FU L H.Monitoring lettuce growth using K-meanscolorimagesegmentation andprincipalcomponent analysis method.Transactions ofthe Chinese Society of Agricultural Engineering,2016,32(12):179-186.(in English)

[10]張洪超,侯德文.一種基于HIS顏色空間的分割復(fù)雜背景圖像算法.山東師范大學(xué)學(xué)報(bào)(自然科學(xué)版),2015,30(3):49-52.ZHANG H C,HOU D W.An image segmentation algorithm with complex background based on his color space.Journal of Shandong Normal University（Natural Science),2015,30(3):49-52.(in Chinese with English abstract)

[11]通霏,武佩,韓丁,等.基于顏色特征的牧草圖像分割方法研究.農(nóng)機(jī)化研究,2014(5):43-47.TONG F,WU P,HAN D,et al.Study on the image segmentation of forage using color features.Journal of Agricultural Mechanization Research,2014(5):43-47.(in Chinese with English abstract)

[12] 陳毅,劉曉玉,蔣崢,等.基于RGB和HSV的膠囊異囊缺陷識(shí)別方法.計(jì)算機(jī)工程與設(shè)計(jì),2014,35(11):3888-3892.CHEN Y,LIU X Y,JIANG Z,et al.Capsule Yinang defect recognition based on RGB and HSV color space.Computer Engineering and Design,2014,35(11):3888-3892.(in Chinese with English abstract)

[13]MORA M,AVILA F,CARRASCO-BENAVIDES M,et al.Automated computation of leaf area index from fruit trees using improved image processing algorithms applied to canopy cover digital photograpies.Computers and Electronics in Agriculture,2016,123:195-202.

[14]張武,黃帥,汪京京,等.復(fù)雜背景下小麥葉部病害圖像分割方法研究.計(jì)算機(jī)工程與科學(xué),2015,37(7):1349-1354.ZHANG W,HUANG S,WANG J J,et al.A segmentation method for wheat leaf images with disease in complex background.Computer Engineering&Science,2015,37(7):1349-1354.(in Chinese with English abstract)

[15]CHITADE A Z,KATIYAR D S K.Colour based image segmentation using K-means clustering.International Journal of Engineering Science and Technology,2010,2(10):5319-5325.

[16]BARBEDO J G A.A novel algorithm for semi-automatic segmentation of plant leaf disease symptoms using digital image processing.Tropical Plant Pathology,2016,41(4):210-224.

[17]LU H,CAO Z G,XIAO Y,et al.Region-based colour modelling forjointcrop and maize tasselsegmentation.Biosystems Engineering,2016,147:139-150.

[18]PREETI,AHUJA K.Colour image segmentation using K-means,fuzzy C-means and density based clustering.International Journal for Research in Applied Science and Engineering Technology(Ijraset),2014,2(VI):31-35.

[19]李小琦.基于Matlab的圖像閾值分割算法研究.軟件導(dǎo)刊,2014,13(12):76-78.LI X Q.Research on image threshold segmentation algorithm based on Matlab.Software Guide,2014,13(12):76-78.(in Chinese)

[20]LIU Y,HAN J,LIU D D,et al.Effect of plastic film mulching on the grain filling and hormonal changes of maize under different irrigation conditions.PLoS One,2015,10(4):1-17.

[21]張仙梅,黃高寶,李玲玲,等.覆膜方式對(duì)旱作玉米硝態(tài)氮時(shí)空動(dòng)態(tài)及氮素利用效率的影響.干旱地區(qū)農(nóng)業(yè)研究,2011,29(5):26-32.ZHANG X M,HUANG G B,LI L L,et al.Effects of mulching patterns on spatio-temporal variation of soil nitrate and nitrogen utilization efficiency of maize on dry land.Agricultural Research in the Arid Areas,2011,29(5):26-32.(in Chinese with English abstract)

[22]SELIM S Z,ISMAIL M A.K-means-type algorithms:A generalized convergence theorem and characterization of local optimality.IEEE Transactions on Pattern Analysis and Machine Intelligence,1984,6(1):81-87.

[23]WOEBBECKE D M,MEYER G E,VON BARGEN K,et al.Color indices for weed identification under various soil,residue,and lighting conditions.Transactions of the ASAE,1995,38(1):259-269.

[24]PéREZ A J,LóPEZ F,BENLLOCH J V,et al.Colour and shape analysis techniques for weed detection in cereal fields.Computers and Electronics in Agriculture,2000,25(3):197-212.

[25]MEYER G E,NETO J C.Verification of color vegetation indices for automated crop imaging applications.Computers and Electronics in Agriculture,2008,63(2):282-293.

[26]OTSU N.A threshold selection method from gray-level histograms.IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1):62-66.

[27]ABEDINPOUR M,SARANGI A,RAJPUT T B S,et al.Performance evaluation of AquaCrop model for maize crop in a semi-arid environment.Agricultural Water Management,2012,110(3):55-66.

[28]BOOTH D T,COX S E,BERRYMAN R D.Point sampling digital imagery with “Samplepoint”.Environmental Monitoring and Assessment,2006,123(1/2/3):97-108.

Image segmentation method of plastic-film corn canopy.Journal of Zhejiang University(Agric.&Life Sci.),2017,43(5):649-656

ZHANG Wanhong,LIU Wenzhao*
(Institute of Soil and Water Conservation,Northwest A&F University,Yangling 712100,Shaanxi,China)

hue;saturation;K-mean clustering analysis;image segmentation;corn canopy

S 513；TP 391

A

10.3785/j.issn.1008-9209.2017.01.071

Summary Percent ground cover of vegetation is an important parameter which

attention of both agronomists and ecologists.Not only does it reflect dynamic growth of plants in a long time,but also it is associated with abstraction of photosynthesis available radiation(APAR)of plants.So far as the maize crop cover is concerned,current researches mainly focused on calculating percent ground cover of maize on bare ground.It is a fact that plastic film mulching has been widely adopted for maize planting due to its effect on reducing water loss,regulating soil temperature,improving the infiltration of rainwater into the soil,enhancing soil water retention,accelerating crop growth,and significantly increasing crop yield.In addition,the recent advances in image analysis software offered potential for analyzing the digital camera images of habitat to objectively quantify ground cover of vegetation in a repeatable and timely manner too.Here we evaluated use of Matlab software for analyzing the digital photographs of plastic-film maize to quantify the percent ground cover.

In this study,the images of plastic-film maize were firstly taken by smart phone under weak light condition,which were JPEG(joint photographic expert group)format here and were in 1 358×1 314 resolution.Then the method combined the K-mean clustering analysis of hue(H)and saturation(S)color components with performing a corresponding mathematical operation was proposed to discriminate the maize and background.The proposed method was comprised of three main steps.First,color images yielding red(R),green(G),and blue(B)subimages were mathematically transformed to hue(H),saturation(S),and intensity(I)ones.And then,the images were respectively segmented using the methods of excess green(ExG),excess red(ExR),excess green minus excess red(ExG－ExR),and Otsu thresholding of excess green,excess red and excess green minus excess red.Second,the K-mean clustering analysis of H and S color components was carried out.Finally,the color difference operation between the K-mean clustering analysis of H and S color components was performed for segmentation of target object.

國(guó)家高技術(shù)研究發(fā)展計(jì)劃（863計(jì)劃）（2013AA102904）。

劉文兆（http://orcid.org/0000-0002-7798-8235），E-mail:wzliu@ms.iswc.ac.cn

（First author）：張萬(wàn)紅（http://orcid.org/0000-0002-0101-3220），E-mail:zhwhong@nwafu.edu.cn

2017-01-07；接受日期（Accepted）：2017-05-09

Results of images processing indicated that the images,which were segmented respectively by excess green,excess red,excess green minus excess red,and Otsu thresholding of excess green,excess red and excess green minus excess red,showed incomplete construct of maize and plastic film,but relatively satisfactory results were achieved by clustering analysis of H and S color components.Specifically,the K-mean clustering analysis of H color component clearly delineated leaf edge of maize,and the K-mean clustering analysis of S color component produced complete plastic film construct.The maize plant was successfully separated from plastic film,soil and other backgrounds by application of the color difference operation between the K-mean clustering analysis of H and S color components.Root mean square error(RMSE)and error rate were calculated to verify the reliability of the method proposed in this paper for segmentation of maize plant.The results showed that the RMSE and error rate of segmentation were 0.004 2 and 3.37%,respectively.The low RMSE and error rate further confirmed the rationality of the method used in this paper.

In conclusion,the method presented in this paper for image segmentation of plastic-film corn canopy is reliable under the weak light condition.