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

?

基于無(wú)人機(jī)影像的可見(jiàn)光波段植被信息識(shí)別

2020-04-09 06:21高永剛林悅歡溫小樂(lè)簡(jiǎn)文彬龔應(yīng)雙
關(guān)鍵詞:植被指數(shù)直方圖波段

高永剛,林悅歡,溫小樂(lè),簡(jiǎn)文彬,2,龔應(yīng)雙

基于無(wú)人機(jī)影像的可見(jiàn)光波段植被信息識(shí)別

高永剛1,2,3,林悅歡1,溫小樂(lè)1※,簡(jiǎn)文彬1,2,龔應(yīng)雙3

(1. 福州大學(xué)環(huán)境與資源學(xué)院,福州 350116;2. 地質(zhì)工程福建省高校工程研究中心,福州 350116;3. 數(shù)字中國(guó)研究院(福建),福州 350108)

該文通過(guò)對(duì)6種典型地物在無(wú)人機(jī)影像可見(jiàn)光波段的光譜特性分析,提出一種基于紅、綠、藍(lán)波段的可見(jiàn)光植被指數(shù)—超綠紅藍(lán)差分指數(shù)EGRBDI(excess green-red-blue difference index),并運(yùn)用該植被指數(shù)與18種基于可見(jiàn)光波段的植被指數(shù)進(jìn)行精度比較研究。研究表明,在利用均值和1倍標(biāo)準(zhǔn)差獲得的區(qū)間范圍內(nèi),EGRBDI各地類(lèi)之間的信息無(wú)重疊交叉現(xiàn)象;該指數(shù)能對(duì)植被覆蓋相對(duì)稀疏區(qū)域進(jìn)行植被信息識(shí)別,其總體精度為97.67%,Kappa系數(shù)為0.941 5,較其他18種指數(shù)具有更好的植被信息識(shí)別能力。利用不同地物覆蓋情況的3幅無(wú)人機(jī)影像作為數(shù)據(jù)源,對(duì)EGRBDI適用性和穩(wěn)定性進(jìn)行研究,結(jié)果表明,在3個(gè)研究區(qū)中,基于EGRBDI的植被信息識(shí)別總精度均高于93%,Kappa系數(shù)均大于0.85,提取精度受地物類(lèi)型差異影響的波動(dòng)性較小,能較好地削弱影像中陰影等因素的影響,具有較好的適用性、可靠性和提取精度。

遙感;植被;光譜分析;無(wú)人機(jī);可見(jiàn)光波段;超綠紅藍(lán)差分指數(shù)

0 引 言

無(wú)人機(jī)(unmanned aerial vehicle, UAV)作為一種有動(dòng)力、可控制、能攜帶多種任務(wù)設(shè)備、執(zhí)行多種任務(wù)的低空飛行平臺(tái)[1-2],其與遙感技術(shù)相結(jié)合具有成本低、操作簡(jiǎn)單、影像獲取速度快和高空間分辨率等傳統(tǒng)衛(wèi)星遙感技術(shù)所無(wú)可比擬的優(yōu)勢(shì)[3-4],已被廣泛用于國(guó)土監(jiān)測(cè)與城市管理、地質(zhì)災(zāi)害、環(huán)境監(jiān)測(cè)以及應(yīng)急保障等重要領(lǐng)域[5-11]。

在遙感領(lǐng)域中,植被指數(shù)是對(duì)地表狀況簡(jiǎn)單有效的度量手段,可以有效地反映植被活力與植被信息[12]?;跓o(wú)人機(jī)影像的可見(jiàn)光植被指數(shù)已被用于農(nóng)作物識(shí)別[13-14]、生物量估算[15-17]和信息提取[18-19]等研究。井然等[20]選用一組最優(yōu)可見(jiàn)光植被指數(shù),結(jié)合面向?qū)ο蟮姆椒▉?lái)提取水生植被;毛智慧等[21]通過(guò)將RGB彩色空間轉(zhuǎn)換為HSL彩色空間,提出一種歸一化色調(diào)亮度植被指數(shù)NHLVI,并將其與超綠指數(shù)ExG、超紅超綠差分指數(shù)ExGR、歸一化綠紅差分指數(shù)NGRDI等7種可見(jiàn)光指數(shù)和實(shí)測(cè)的NDVI數(shù)據(jù)進(jìn)行相關(guān)性比較試驗(yàn)和ROC曲線(xiàn)分析,證明NHLVI與實(shí)測(cè)NDVI數(shù)據(jù)具有較高相關(guān)性且提取植被精度較高;滕佳昆等[22]運(yùn)用超綠指數(shù)ExG、相對(duì)綠度指數(shù)Gcc、綠紅植被指數(shù)GRVI和色相指數(shù)對(duì)黃土高原上不同生長(zhǎng)時(shí)間節(jié)點(diǎn)的刺槐RGB圖像進(jìn)行試驗(yàn)對(duì)比,試驗(yàn)表明,基于平均值法的ExG指數(shù)識(shí)別SOS和EOS時(shí)期的刺槐最接近實(shí)測(cè)值;吳蘭蘭等[23]運(yùn)用ExG、ExGR和COM等6種可見(jiàn)光植被指數(shù)和Otsu算法分割含陰影區(qū)域的大田油菜圖像,運(yùn)用ROC法定量評(píng)價(jià)6種指數(shù)的分割效果,試驗(yàn)結(jié)果表明,COM指數(shù)分割結(jié)果最佳;Kazmi等[24]運(yùn)用ExG、CIVE、GB等14種可見(jiàn)光植被指數(shù)針對(duì)甜菜田中的薊進(jìn)行檢測(cè),其中ExG、GB和CIVE的平均準(zhǔn)確率均達(dá)90%以上;Liu等[25]建立一種基于無(wú)人機(jī)影像評(píng)價(jià)小麥出苗率均勻性的綜合方法,其結(jié)合ExG指數(shù)和Otsu法進(jìn)行小麥幼苗的覆蓋范圍研究;Wan等[26]基于開(kāi)花期的油菜花RGB影像,統(tǒng)計(jì)分析8種常用可見(jiàn)光植被指數(shù)和2種多光譜植被指數(shù)與油菜花數(shù)之間的相關(guān)關(guān)系,證明結(jié)合基于無(wú)人機(jī)RGB影像的植被指數(shù)分類(lèi)圖像,對(duì)估算油菜花數(shù)具有巨大潛力;Zhang等[27]統(tǒng)計(jì)分析9種可見(jiàn)光植被指數(shù)與甘蔗、玉米、棉花和水稻4種植被覆蓋度之間的關(guān)系,發(fā)現(xiàn)不同作物類(lèi)型對(duì)應(yīng)的植被覆蓋度模型準(zhǔn)確性和最佳植被指數(shù)也不同,其中 CIVE指數(shù)對(duì)于4種作物混合種植養(yǎng)殖區(qū)域的植被覆蓋度估算效果最佳。目前,常用的植被指數(shù)多為基于可見(jiàn)光—近紅外波段構(gòu)建,單純基于可見(jiàn)光波段構(gòu)建的植被指數(shù)相對(duì)較少,使得構(gòu)建一種具有普適性且適用于無(wú)人機(jī)可見(jiàn)光波段的植被指數(shù)顯得十分必要。

本研究在分析僅含可見(jiàn)光波段的無(wú)人機(jī)影像的各地物間光譜曲線(xiàn)特點(diǎn)的基礎(chǔ)上,提出了一種能夠有效進(jìn)行植被信息識(shí)別的超綠紅藍(lán)差分指數(shù)(EGRBDI),并與18種常用的可見(jiàn)光植被指數(shù)進(jìn)行對(duì)比分析,以期為無(wú)人機(jī)在植被指數(shù)創(chuàng)建和植被信息識(shí)別方面的應(yīng)用研究提供參考。

1 基本原理與研究方法

1.1 超綠紅藍(lán)差分指數(shù)(EGRBDI)指數(shù)

本研究選用地物類(lèi)別豐富且具有較好區(qū)域代表性的一幅拍攝于福建省三明市寧化縣翠江鎮(zhèn)的無(wú)人機(jī)可見(jiàn)光波段影像數(shù)據(jù)作為數(shù)據(jù)源進(jìn)行研究(圖1和表1),影像拍攝時(shí)天氣良好,所獲得的無(wú)人機(jī)影像受氣象等因素的影響較小。由于本研究?jī)?nèi)容不涉及各個(gè)波段的中心波長(zhǎng)位置和波段的范圍,故所獲得的無(wú)人機(jī)影像不必進(jìn)行嚴(yán)格的輻射校正。

圖1 研究區(qū)域無(wú)人機(jī)影像

表1 數(shù)據(jù)源主要參數(shù)

由于目前常用的可見(jiàn)光植被指數(shù)易受地物類(lèi)型差異的影響且可靠性較差,在對(duì)研究區(qū)內(nèi)無(wú)人機(jī)影像地物特征分析的基礎(chǔ)上,對(duì)可見(jiàn)光波段植被指數(shù)進(jìn)行改進(jìn)研究。為了使所構(gòu)建的植被指數(shù)具有更好的通用性和可靠性,采用以下地物類(lèi)別樣區(qū)的確定原則:(1)對(duì)于每一類(lèi)地類(lèi),選擇面積適當(dāng)且與邊界具有一定距離的同質(zhì)區(qū)域;(2)選取的每類(lèi)地物應(yīng)覆蓋各種亮度區(qū)域,即包含低、中、高亮度區(qū)域,最大程度地覆蓋不同亮度的動(dòng)態(tài)范圍,確保在較寬的亮度值范圍分析各地物之間的關(guān)系;(3)每一類(lèi)地物選擇的樣區(qū)數(shù)目盡量均衡,對(duì)于某些色澤和亮度差異較大的地類(lèi)選擇樣區(qū)的數(shù)目可以適度增加。

根據(jù)上述3個(gè)原則,將圖1中的地物分為樹(shù)木、草地、農(nóng)田、水泥路、裸土和建筑物6種類(lèi)型,并進(jìn)行植被與非植被區(qū)域在可見(jiàn)光波段的光譜特性分析;然后,根據(jù)分析獲得的各地類(lèi)在不同波段間像元值差異性構(gòu)建一種新的基于可見(jiàn)光的植被指數(shù)。本研究運(yùn)用人機(jī)交互的方式通過(guò)ENVI軟件人工勾畫(huà)出200個(gè)代表樣區(qū)并進(jìn)行野外實(shí)地調(diào)查檢驗(yàn),其中樹(shù)木30、草地30、農(nóng)田50、水泥地30、裸土30、建筑物30。利用均值作為各類(lèi)地物在可見(jiàn)光波段之間像元值總體差異的評(píng)價(jià)指標(biāo),以標(biāo)準(zhǔn)差對(duì)各類(lèi)地物在各波段中像元值波動(dòng)范圍進(jìn)行評(píng)價(jià)(表2);并以1倍標(biāo)準(zhǔn)差區(qū)間范圍評(píng)價(jià)各地類(lèi)之間的信息是否存在重疊交叉的現(xiàn)象(圖2)。

表2 不同地物類(lèi)型在紅、綠、藍(lán)波段的像元統(tǒng)計(jì)值

圖2 不同地物類(lèi)型在藍(lán)、綠、紅波段的均值和1倍標(biāo)準(zhǔn)差區(qū)間范圍

通過(guò)對(duì)表2和圖2分析可知,各地類(lèi)在紅(R)、綠(G)、藍(lán)(B)波段間的變化趨勢(shì)與其地物反射波譜曲線(xiàn)變化趨勢(shì)基本相吻合,其中植被類(lèi)別(樹(shù)木、草地和農(nóng)田)的變化趨勢(shì)與典型健康植被光譜曲線(xiàn)相一致。樹(shù)木、草地和農(nóng)田像元均值呈現(xiàn)出先升后降的變化趨勢(shì),水泥路和裸土兩類(lèi)地物像元均值呈現(xiàn)出遞增趨勢(shì),而建筑物像元均值呈現(xiàn)出遞減的趨勢(shì),除建筑物在綠光波段和紅光波段與草地有部分重合外,植被類(lèi)別和非植被類(lèi)別的之間的數(shù)值范圍在R、G、B波段無(wú)明顯重疊,這說(shuō)明藍(lán)光波段具有將建筑物及草地區(qū)分開(kāi)的優(yōu)勢(shì)。綜上所述,對(duì)于植被與非植被信息的識(shí)別不能依靠單一波段,應(yīng)綜合考慮不同地物類(lèi)別在R、G、B波段上的光譜特性,可以起到增大植被與非植被信息差異的效果。

為了實(shí)現(xiàn)無(wú)人機(jī)影像對(duì)植被和非植被信息的有效識(shí)別,通過(guò)利用RGBVI指數(shù)進(jìn)行大量實(shí)驗(yàn)研究分析可知RGBVI指數(shù)對(duì)植被信息識(shí)別具有較高精度,但其在植被稀疏區(qū)域提取植被信息能力較弱,故借鑒RGBVI的構(gòu)建原理,在綜合考慮圖1中6種地物類(lèi)型的RGB波段光譜特性規(guī)律的基礎(chǔ)上,構(gòu)建一種基于可見(jiàn)光波段的植被指數(shù),以進(jìn)一步優(yōu)化植被信息的提取能力。由于健康綠色植被在綠光波段有強(qiáng)反射,在藍(lán)光和紅光波段有強(qiáng)吸收,所以通過(guò)利用2倍的綠波段的平方進(jìn)一步增強(qiáng)植被在綠光波段的強(qiáng)反射作用;非植被類(lèi)別的綠光波段分別和藍(lán)光波段和紅光波段的差值均比植被小,利用2倍的綠波段的平方減去紅、藍(lán)2個(gè)可見(jiàn)光波段的乘積,使植被類(lèi)別信息識(shí)別范圍相對(duì)擴(kuò)大,非植被類(lèi)別信息識(shí)別范圍相對(duì)縮小,綜合兩方面的作用,使植被類(lèi)別信息和非植被類(lèi)別信息之間無(wú)重疊部分(圖3),更有利于植被信息的識(shí)別。構(gòu)建的基于可見(jiàn)光波段的植被指數(shù)EGRBDI(excess green-red-blue difference index),公式如下:

式中R、G、B分別代表影像中的像元亮度值或反射率,EGRBDI的數(shù)值范圍為[-1,1]。

1.2 基于閾值的無(wú)人機(jī)影像植被信息識(shí)別

利用植被指數(shù)進(jìn)行植被信息的識(shí)別,關(guān)鍵是設(shè)定合適的閾值對(duì)影像中的植被與非植被信息進(jìn)行區(qū)分,在植被指數(shù)計(jì)算得到的值中,高于閾值的部分歸為植被,低于閾值的部分歸為非植被,從而達(dá)到識(shí)別圖像中植被信息的目的。因此,無(wú)人機(jī)影像植被信息識(shí)別精度的高低取決于閾值選取的優(yōu)劣。本研究利用雙峰直方圖法和最大熵值法確定各植被指數(shù)識(shí)別植被信息的相應(yīng)閾值,并通過(guò)精度比較法確定各植被指數(shù)最終的閾值。

1)雙峰直方圖法。雙峰直方圖是指圖像的像素灰度值基本集中于2處,即影像的直方圖中包含2個(gè)“山峰”,這2個(gè)最高點(diǎn)位置的灰度值對(duì)應(yīng)于對(duì)象內(nèi)部或外部的典型灰度值,兩峰之間峰谷所對(duì)應(yīng)的值為對(duì)象之間邊緣附近點(diǎn)的位置[28],通常選取兩峰之間的坡谷所對(duì)應(yīng)的值作為閾值。

2)最大熵值法。此方法通過(guò)假設(shè)閾值為,閾值把影像分為目標(biāo)和背景,通過(guò)計(jì)算圖像的累加概率直方圖和各個(gè)灰度級(jí)的熵,最后計(jì)算目標(biāo)區(qū)域的熵H()和背景區(qū)域的熵H(),當(dāng)()=H()+H()取得最大值時(shí),該值對(duì)應(yīng)的即為最佳閾值[29]。

2 試驗(yàn)結(jié)果與分析

2.1 可見(jiàn)光植被指數(shù)的計(jì)算與分析

在遙感領(lǐng)域中,目前存在的植被指數(shù)已有上百種,而基于可見(jiàn)光波段的植被指數(shù)相對(duì)較少(表3)。運(yùn)用Python編程語(yǔ)言根據(jù)表3和式(1)的公式分別計(jì)算數(shù)據(jù)源影像各可見(jiàn)光植被指數(shù),獲得19種植被指數(shù)灰度直方圖。為便于比較不同指數(shù),本研究將各可見(jiàn)光植被指數(shù)的計(jì)算結(jié)果利用極差標(biāo)準(zhǔn)化進(jìn)行歸一化處理,使各指數(shù)計(jì)算結(jié)果圖的數(shù)值范圍限定在[0,1]區(qū)間內(nèi),同時(shí)將RGRI、GBRI、CIVE、ExR和ExB這5種植被指數(shù)的計(jì)算結(jié)果取反,使得19種可見(jiàn)光植被指數(shù)均以暗色區(qū)代表非植被信息,亮色區(qū)代表植被信息(圖4)。由圖4可知,EGRBDI、GLI、ExG、、CIVE、ExGR、COM、COM2、RGBVI和V-MSAVI這10種可見(jiàn)光植被指數(shù)的植被與非植被區(qū)域的對(duì)比度較為明顯,植被信息識(shí)別效果較好,其中植被區(qū)域?yàn)榱涟咨侵脖粎^(qū)域?yàn)榘岛谏虬祷疑?;而NGRDI、RGRI、GBRI、NGBDI、WI、VEG、ExR、ExB和MGRVI這9種可見(jiàn)光植被指數(shù)的植被與非植被區(qū)域的對(duì)比度較差,其中NGRDI、RGRI、GBRI、NGBDI、VEG、ExR、ExB和MGRVI指數(shù)圖中稀疏植被區(qū)與裸土的灰度值相近,將導(dǎo)致在后續(xù)分類(lèi)過(guò)程中這兩種地物容易出現(xiàn)混淆現(xiàn)象。

表3 可見(jiàn)光植被指數(shù)

續(xù)表

圖4 各可見(jiàn)光植被指數(shù)計(jì)算結(jié)果

為更好地對(duì)比19種可見(jiàn)光植被指數(shù)的提取結(jié)果,本研究統(tǒng)計(jì)各植被指數(shù)在6類(lèi)典型地物ROI區(qū)域的統(tǒng)計(jì)特征值(表4)。由表4可知,EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI在植被信息和非植被信息之間無(wú)重疊交叉部分,對(duì)于僅利用可見(jiàn)光3個(gè)波段進(jìn)行植被信息識(shí)別具有較好的效果;而NGRDI、RGRI、GBRI、NGBDI、WI、VEG、ExR、ExGR、COM、ExB、COM2和MGRVI這12種可見(jiàn)光植被指數(shù)在植被信息和非植被信息之間具有不同程度的重疊交叉部分,因此在植被信息識(shí)別過(guò)程中,會(huì)造成植被與非植被區(qū)域存在不同程度的誤分或漏分現(xiàn)象。因此,基于無(wú)人機(jī)影像進(jìn)行植被信息識(shí)別時(shí),應(yīng)優(yōu)先從EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI這7種植被指數(shù)中選取合適的指數(shù)。

表4 基于ROI的19種可見(jiàn)光植被指數(shù)的統(tǒng)計(jì)值

2.2 植被信息識(shí)別與精度評(píng)價(jià)

為確定19種可見(jiàn)光植被指數(shù)進(jìn)行植被信息識(shí)別時(shí)的閾值,本研究統(tǒng)計(jì)各可見(jiàn)光植被指數(shù)對(duì)應(yīng)的灰度直方圖,并將19種指數(shù)的灰度直方圖歸一化至[0, 255]區(qū)間以便各指數(shù)間的比較分析(圖5)。對(duì)于灰度直方圖存在明顯雙峰的指數(shù),其對(duì)應(yīng)的地物區(qū)分性能相對(duì)較強(qiáng)。由圖5分析可知,WI指數(shù)幾乎無(wú)雙峰特征,VEG和COM2指數(shù)雙峰特征不明顯且兩峰間的距離很近;EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI指數(shù)有明顯的雙峰特征且直方圖沒(méi)有過(guò)多的刺峰;其他9種指數(shù)的灰度直方圖雖有明顯的雙峰特征,但其雙峰存在不同程度的刺峰現(xiàn)象或雙峰間距離相鄰很近的情況,故EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI指數(shù)對(duì)于植被識(shí)別的能力相對(duì)于其他12種指數(shù)較好。

圖5 19種可將光植被指數(shù)統(tǒng)計(jì)直方圖

利用雙峰直方圖法和最大熵值法分別計(jì)算19種可見(jiàn)光植被指數(shù)的閾值,并利用參考影像運(yùn)用精度比較法對(duì)比兩種方法得到的分類(lèi)精度,以精度較大者所對(duì)應(yīng)的閾值確定為各植被指數(shù)的最終閾值。為削弱人為因素干擾,客觀地對(duì)比各植被指數(shù)的分類(lèi)精度,采用隨機(jī)森林的分類(lèi)方法在ENVI軟件中生成植被與非植被區(qū)域分類(lèi)參考圖(圖6a)。利用參考影像進(jìn)行分類(lèi)精度定量評(píng)價(jià),經(jīng)比較分析可知,除GBRI和ExB利用最大熵值法計(jì)算獲得的分類(lèi)精度比雙峰直方圖法高,WI和VEG在2種閾值確定方法的精度一致外,其他15種可見(jiàn)光植被指數(shù)均在雙峰直方圖法中獲得較大精度,因此,本研究利用最大熵法確定GBRI和ExB的閾值,利用雙峰直方圖法確定其他17種可見(jiàn)光植被指數(shù)的閾值(表5)。

根據(jù)表5所確定的閾值進(jìn)行植被信息識(shí)別,得到19種植被指數(shù)對(duì)應(yīng)的植被分類(lèi)結(jié)果(圖6b~6t)。利用分類(lèi)參考圖(圖6a)評(píng)價(jià)19種植被指數(shù)識(shí)別植被信息的精度,得到各植被指數(shù)提取精度的定量評(píng)價(jià)結(jié)果(表5)。通過(guò)對(duì)圖6分析可知,NGRDI、GBRI、WI、VEG和ExB指數(shù)在高植被覆蓋度區(qū)域識(shí)別植被信息效果較差;在植被稀疏區(qū)域,EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI指數(shù)的提取效果較好,另12種指數(shù)存在不同程度的漏分或錯(cuò)分現(xiàn)象;在建筑物區(qū)域,19種指數(shù)同時(shí)存在不同程度的錯(cuò)分現(xiàn)象,主要原因是建筑物部分區(qū)域含有色調(diào)與植被相似的綠色防護(hù)網(wǎng)和部分陰影區(qū)域,容易導(dǎo)致對(duì)植被信息敏感的指數(shù)將其錯(cuò)分為植被和對(duì)陰影敏感的指數(shù)造成錯(cuò)分現(xiàn)象,其中以NGRDI、RGRI、ExR和MGRVI指數(shù)錯(cuò)分現(xiàn)象最為嚴(yán)重;在裸土區(qū)域,可以明顯看出NGBDI指數(shù)存在嚴(yán)重地錯(cuò)分現(xiàn)象,如影像左下部分裸土道路均被錯(cuò)分為植被。綜上所述,通過(guò)定性分析可知,19種植被指數(shù)中以EGRBDI、GLI、ExG、、CIVE、RGBVI和V-MSAVI這7種指數(shù)識(shí)別植被信息效果較好,所對(duì)應(yīng)的分類(lèi)結(jié)果圖與參考分類(lèi)圖基本一致。

注:綠色為植被區(qū)域,灰色為非植被區(qū)域。

表5 19種可見(jiàn)光植被指數(shù)的植被提取精度評(píng)價(jià)

根據(jù)表5的定量評(píng)價(jià)結(jié)果可知,EGRBDI識(shí)別植被信息的總精度為97.67%,Kappa系數(shù)為0.941 5,其精度優(yōu)于其他18種可見(jiàn)光植被指數(shù),且EGRBDI在植被和非植被的識(shí)別正確率也較高。由圖5和表5可知,對(duì)于雙峰特征不明顯或雙峰有刺峰現(xiàn)象的植被指數(shù),其對(duì)應(yīng)的總體精度和Kappa系數(shù)也相應(yīng)較低,兩者存在很強(qiáng)的相關(guān)性,即指數(shù)提取結(jié)果的直方圖雙峰形狀對(duì)該指數(shù)最終識(shí)別植被信息的效果具有直接影響。對(duì)比表4和表5可以發(fā)現(xiàn),表5中前7種植被指數(shù)提取植被信息的精度較高,其對(duì)應(yīng)的6種典型地類(lèi)ROI區(qū)域的植被與非植被信息之間無(wú)重疊交叉區(qū)域,另外12種精度較低的植被指數(shù)對(duì)應(yīng)的植被與非植被信息之間存在不同度的重疊交叉部分。

為進(jìn)一步比較7種精度較高指數(shù),本研究從細(xì)節(jié)部分比較其在圖6a紅色圈區(qū)域A中的植被信息提取結(jié)果(圖7),從圖7可以看出,EGRBDI對(duì)于區(qū)域A的稀疏植被區(qū)域提取效果比其他6種指數(shù)好,由此可見(jiàn),在植被覆蓋度較低的情況下,EGRBDI指數(shù)有較大的優(yōu)勢(shì)。

圖7 區(qū)域A對(duì)應(yīng)的7種指數(shù)植被信息提取結(jié)果

2.3 適用性評(píng)價(jià)

為了更好地驗(yàn)證EGRBDI可見(jiàn)光植被指數(shù)的適用性和準(zhǔn)確性,選擇3幅無(wú)人機(jī)影像利用相同的方法對(duì)數(shù)據(jù)源中精度較高的前5種植被指數(shù)(EGRBDI、CIVE、RGBVI、GLI和V-MSAVI)進(jìn)行植被信息提取。3幅影像均與實(shí)驗(yàn)影像同一時(shí)間段在同一地點(diǎn)附近拍攝,其中研究區(qū)1(圖8a)含有布滿(mǎn)綠藻的池塘,拍攝高度為500 m;研究區(qū)2(圖8g)含有渾濁河流,拍攝高度為500 m;研究區(qū)3(圖8m)含有存在明顯的陰影區(qū)域,且圖像中部的河流顏色較綠,拍攝高度為700 m。3個(gè)研究區(qū)的5種植被指數(shù)利用雙峰直方圖法確定的閾值所識(shí)別植被信息的結(jié)果圖如圖8b~8f、圖8h~8l和圖8n~8r所示。通過(guò)對(duì)圖8的分析可知,EGRBDI、GLI和V-MSAVI植被信息提取效果在研究區(qū)1(圖8a)中獲得較佳的效果,并且這3種指數(shù)能有效地抑制左上方部分藍(lán)色屋頂信息和區(qū)分右上方顏色偏綠的小池塘水體信息,而其他2種指數(shù)在相應(yīng)位置會(huì)存在較為嚴(yán)重的錯(cuò)分現(xiàn)象;在研究區(qū)2(圖8g)中,CIVE存在將少部分建筑物信息錯(cuò)分為植被信息的情況,而其他4種植被指數(shù)都能得到較好的提取效果;在研究區(qū)3(圖8m)中,EGRBDI總體植被信息識(shí)別效果較好,能有效抑制陰影信息、建筑物信息和顏色偏綠的水體信息,不會(huì)過(guò)多地造成椒鹽現(xiàn)象和錯(cuò)分現(xiàn)象,另外4種指數(shù)在中部顏色偏綠的水體區(qū)域存在不同程度地錯(cuò)分現(xiàn)象,同時(shí)在不同的建筑物和陰影區(qū)域出現(xiàn)的椒鹽現(xiàn)象比EGRBDI嚴(yán)重。綜上所述,EGRBDI在3個(gè)研究區(qū)均獲得較好的植被信息識(shí)別效果,相對(duì)具有較好的適用性和穩(wěn)定性。

為了進(jìn)一步評(píng)價(jià)EGRBDI和其他4種指數(shù)的植被信息識(shí)別精度,采用隨機(jī)抽樣的方法,在3幅影像中各隨機(jī)布置400個(gè)點(diǎn),并利用人機(jī)交互的方式統(tǒng)計(jì)出各研究區(qū)5種指數(shù)的精度評(píng)價(jià)數(shù)據(jù)以進(jìn)行定量評(píng)價(jià)(表6)。由表6可知,各研究區(qū)5種指數(shù)均獲得較為理想的精度, 5種指數(shù)在不同研究區(qū)中的精度差異大小各不相同。5種植被指數(shù)在研究區(qū)2中總精度和Kappa系數(shù)之間的差異相對(duì)較小,而在研究區(qū)1和研究區(qū)3中對(duì)應(yīng)的總精度和Kappa系數(shù)之間的差異相對(duì)較大。EGRBDI在不同研究區(qū)均能獲得相對(duì)較優(yōu)的結(jié)果,其植被和非植被正確率均高于90%,總精度均高于93%,Kappa系數(shù)均大于0.85,精度波動(dòng)性較小。綜上所述,EGRBDI的適用性強(qiáng)、準(zhǔn)確性高,能夠有效準(zhǔn)確地識(shí)別無(wú)人機(jī)影像中的植被信息,同時(shí),能有效地抑制陰影影響,具有區(qū)分顏色與植被相似的部分水域信息的能力。

圖8 前5種精度較高指數(shù)的精度驗(yàn)證結(jié)果

表6 前5種精度較高指數(shù)的精度驗(yàn)證數(shù)據(jù)

3 結(jié) 論

本研究針對(duì)無(wú)人機(jī)拍攝的高空間分辨率可見(jiàn)光影像,在綜合分析6種典型地類(lèi)在可見(jiàn)光的光譜特性的基礎(chǔ)上,提出了一種基于RGB的可見(jiàn)光植被指數(shù)——超綠紅藍(lán)差分指數(shù)(EGRBDI),并與18種常見(jiàn)的可見(jiàn)光植被指數(shù)進(jìn)行比較研究,同時(shí)選取研究試驗(yàn)中精度較高的前5種植被指數(shù)對(duì)EGEBDI進(jìn)行適用性評(píng)價(jià)分析,試驗(yàn)結(jié)果表明:

1)在利用均值和1倍標(biāo)準(zhǔn)差獲得的區(qū)間范圍內(nèi),EGRBDI各地類(lèi)之間的信息無(wú)重疊交叉現(xiàn)象;其總精度為97.67%,Kappa系數(shù)為0.941 5,植被信息識(shí)別精度不同程度優(yōu)于其他18種可見(jiàn)光植被指數(shù);并在植被覆蓋相對(duì)稀疏的區(qū)域具有更好的植被信息識(shí)別能力;

2)在適用性評(píng)價(jià)試驗(yàn)中,EGRBDI指數(shù)在3個(gè)研究區(qū)的植被提取效果較佳且其精度波動(dòng)性較小,并能有效地抑制陰影信息和削弱與植被色彩相似地水體信息的干擾能力;其他4種指數(shù)的提取精度受地物類(lèi)型差異性的顯著、穩(wěn)定性差,且對(duì)陰影無(wú)削弱能力;但當(dāng)影像中僅有植被和裸土2種地類(lèi)且植被稀疏區(qū)域面積較大的情況時(shí),EGRBDI指數(shù)的植被信息識(shí)別的效果較差。EGRBDI指數(shù)對(duì)于絕大多數(shù)情況下的無(wú)人機(jī)影像具有適用性強(qiáng),準(zhǔn)確性高的優(yōu)點(diǎn)。

[1]Watts A C, Ambrosia V G, Hinkley E A. Unmanned aircraft systems in remote sensing and scientific research: Classification and considerations of use[J]. Remote Sensing, 2012, 4(6): 1671—1692.

[2]紀(jì)景純,趙原,鄒曉娟,等. 無(wú)人機(jī)遙感在農(nóng)田信息監(jiān)測(cè)中的應(yīng)用進(jìn)展[J]. 土壤學(xué)報(bào),2019,56(4):773-784.

Ji Jingchun, Zhao Yuan, Zou Xiaojuan, et al. Advancement in application of UAV remote sensing to monitoring of farmlands[J]. Acta Pedologica Sinica, 2019, 56(4): 773-784. (in Chinese with English abstract)

[3]Mohamed H, Zahra L, Naser E S. A new vegetation segmentation approach for cropped fields based on threshold detection from hue histograms[J]. Sensors, 2018, 18(4): 1253.

[4]王利民,劉佳,楊玲波,等. 基于無(wú)人機(jī)影像的農(nóng)情遙感監(jiān)測(cè)應(yīng)用[J]. 農(nóng)業(yè)工程學(xué)報(bào),2013,29(18):136-145.

Wang Limin, Liu Jia, Yang Lingbo, et al. Applications of unmanned aerial vehicle images on agricultural remote sensing monitoring[J]. Transaction of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2013, 29(18): 136-145. (in Chinese with English abstract)

[5]宋曉陽(yáng),黃耀歡,董東林,等. 融合數(shù)字表面模型的無(wú)人機(jī)遙感影像城市土地利用分類(lèi)[J]. 地球信息科學(xué)學(xué)報(bào),2018,20(5):703-711.

Song Xiaoyang, Huang Yaohuan, Dong Donglin, et al. Urban land use classification from UAV remote sensing images based on digital surface model[J]. Journal of Geo-information Science, 2018, 20(5): 703-711. (in Chinese with English abstract)

[6]王帥永,唐川,何敬,等. 無(wú)人機(jī)在強(qiáng)震區(qū)地質(zhì)災(zāi)害精細(xì)調(diào)查中的應(yīng)用研究[J]. 工程地質(zhì)學(xué)報(bào),2016,24(4):713-719.

Wang Shuaiyong, Tang Chuan, He Jing, et al. Use of unmanned aerial vehicle for precise investigation of geological hazard in strong seismic zone[J]. Journal of Engineering Geology, 2016, 24(4): 713-719. (in Chinese with English abstract)

[7]Chen Jinhong, Liu Haoting, Zheng Jingchen, et al. Damage degree evaluation of earthquake area using UAV aerial image[J]. International Journal of Aerospace Engineering, 2016, 2016(6): 1-10.

[8]洪運(yùn)富,楊海軍,李營(yíng),等. 水源地污染源無(wú)人機(jī)遙感監(jiān)測(cè)[J]. 中國(guó)環(huán)境監(jiān)測(cè),2015,31(5):163-166.

Hong Yunfu, Yang Haijun, Li Ying, et al. Monitoring of water source using unmanned aerial vehicle remote sensing technology[J]. Environmental Monitoring in China, 2015, 31(5): 163-166. (in Chinese with English abstract)

[9]Langhammer J. UAV monitoring of stream restorations[J]. Hydrology, 2019, 6(2): 29.

[10]雷添杰,李長(zhǎng)春,何孝瑩. 無(wú)人機(jī)航空遙感系統(tǒng)在災(zāi)害應(yīng)急救援中的應(yīng)用[J]. 自然災(zāi)害學(xué)報(bào),2011,20(1):178-183.

Lei Tianjie, Li Changchun, He Xiaoying. Application of aerial remote sensing of pilotless aircraft to disaster emergency rescue[J]. Journal of Natural Disasters, 2011, 20(1): 178-183. (in Chinese with English abstract)

[11]Hart A, Chai P R, Griswold M K, et al. Acceptability and perceived utility of drone technology among emergency medical service responders and incident commanders for mass casualty incident management[J]. American Journal of Disaster Medicine, 2017, 12(4): 261-265.

[12]高永平,康茂東,何明珠,等. 基于無(wú)人機(jī)可見(jiàn)光波段對(duì)荒漠植被覆蓋度提取的研究:以沙坡頭地區(qū)為例[J]. 蘭州大學(xué)學(xué)報(bào):自然科學(xué)版,2018,54(6):770-775.

Gao Yongping, Kang Maodong, He Mingzhu, et al. Extraction of desert vegetation coverage based on visible light band information of unmanned aerial vehicle: A case study of Shapotou region[J]. Journal of Lanzhou University: Natural Sciences, 2018, 54(6): 770-775. (in Chinese with English abstract)

[13]Pourdarbani R, Sabzi S, García-Amicis V M, et al. Automatic classification of chickpea varieties using computer vision techniques[J]. Agronomy, 2019, 9(11): 672. DOI: 10.3390/agrorlomy 9110672.

[14]韓文霆,李廣,苑夢(mèng)嬋,等. 基于無(wú)人機(jī)遙感技術(shù)的玉米種植信息提取方法研究[J]. 農(nóng)業(yè)機(jī)械學(xué)報(bào),2017,48(1):139-147.

Han Wenting, Li Guang, Yuan Mengchan, et al. Extraction method of maize planting information based on UAV remote sensing techonology[J]. Transactions of the Chinese Society for Agricultural Machinery, 2017, 48(1): 139-147. (in Chinese with English abstract)

[15]張正健,李?lèi)?ài)農(nóng),邊金虎,等. 基于無(wú)人機(jī)影像可見(jiàn)光植被指數(shù)的若爾蓋草地地上生物量估算研究[J]. 遙感技術(shù)與應(yīng)用,2016,31(1):51-62.

Zhang Zhengjian, Li Ainong, Bian Jinhu, et al. Estimating aboveground biomass of grassland in Zoige by visible vegetation index derived from unmanned aerial vehicle image[J]. Remote Sensing Technology and Application, 2016, 31(1): 51-62. (in Chinese with English abstract)

[16]Niu Yaxiao, Zhang Liyuan, Zhang Huihui, et al. Estimating above-ground biomass of maize using features derived from UAV-based rgb imagery[J]. Remote Sensing, 2019, 11(11): 1261. DOI: 10.3390/rs 11111261.

[17]Lu Ning, Zhou Jie, Han Zixu, et al. Improved estimation of aboveground biomass in wheat from RGB imagery and point cloud data acquired with a low-cost unmanned aerial vehicle system[J]. Plant Methods, 2019, 15(1): 17. DOI: 10.1186/s 13007-019-0402-3.

[18]Yang Wenzhu, Wang Sile, Zhao Xiaolan, et al. Greenness identification based on HSV decision tree[J]. Information Processing in Agriculture, 2015, 2(3/4): 149-160.

[19]Hu Xiao, Li Xinju. Information extraction of subsided cultivated land in high-groundwater-level coal mines based on unmanned aerial vehicle visible bands[J]. Environmental Earth Sciences, 2019, 78(14): 413. DOI: 10.1107/s/2665-019-8417-7.

[20]井然,鄧?yán)?,趙文吉,等. 基于可見(jiàn)光植被指數(shù)的面向?qū)ο鬂竦厮脖惶崛》椒╗J]. 應(yīng)用生態(tài)學(xué)報(bào),2016,27(5):1427-1436.

Jing Ran, Deng Lei, Zhao Wenji, et al. Object-oriented aquatic vegetation extracting approach based on visible vegetation indices[J]. Chinese Journal of Applied Ecology, 2016, 27(5): 1427-1436. (in Chinese with English abstract)

[21]毛智慧,鄧?yán)?,賀英,等. 利用色調(diào)—亮度彩色分量的可見(jiàn)光植被指數(shù)[J]. 中國(guó)圖象圖形學(xué)報(bào),2017,22(11):1602-1610.

Mao Zhihui, Deng Lei, He Ying, et al. Vegetation index for visible-light true-color image using hue and lightness color channels[J]. Journal of Image and Graphics, 2017, 22(11): 1602-1610. (in Chinese with English abstract)

[22]滕佳昆,劉宇,丁明濤. 基于RGB圖像的刺槐季節(jié)變化監(jiān)測(cè)適用指數(shù)研究[J]. 遙感技術(shù)與應(yīng)用,2018,33(3):476-485.

Teng Jiakun, Liu Yu, Ding Mingtao. The evaluation of efficiency of color metrics in monitoring robiuia pseudoacacia phenology based on RGB images[J]. Remote Sensing Technology and Application, 2018, 33(3): 476-485. (in Chinese with English abstract)

[23]吳蘭蘭,熊利榮,彭輝. 基于RGB植被指數(shù)的大田油菜圖像分割定量評(píng)價(jià)[J]. 華中農(nóng)業(yè)大學(xué)學(xué)報(bào),2019,38(2):115-119.

Wu Lanlan, Xiong Lirong, Peng Hui. Quantitative evaluation of in-field rapeseed image segmentation based on RGB vegetation indices[J]. Journal of Huazhong Agricultural University, 2019, 38(2): 115-119. (in Chinese with English abstract)

[24]Kazmi W, Garcia-Ruiz F J, Nielsen J, et al. Detecting creeping thistle in sugar beet fields using vegetation indices[J]. Computers and Electronics in Agriculture, 2015, 112(2): 10-19.

[25]Liu Tao, Li Rui, Jin Xiuliang, et al. Evaluation of seed emergence uniformity of mechanically sown wheat with UAV RGB imagery[J]. Remote Sensing, 2017, 9(12): 1241. DOI: 10.3390/ys9121241.

[26]Wan Liang, Li Yijian, Cen haiyan, et al. Combining UAV-based vegetation indices and image classification to estimate flower number in oilseed rape[J]. Remote Sensing, 2018, 10(9): 1484. DOI: 10.3390/rs10091484.

[27]Zhang Dongdong, Mansaray L R, Jin Hongwei, et al. A universal estimation model of fractional vegetation cover for different crops based on time series digital photographs[J]. Computers and Electronics in Agriculture, 2018, 151(6): 93-103.

[28]梁華為. 直接從雙峰直方圖確定二值化閾值[J]. 模式識(shí)別與人工智能,2002,15(2):253-256.

Liang Huawei. Direct determination of threshold from bimodal histogram[J]. Pattern Recognition and Artificial Intelligence, 2002, 15(2): 253-256. (in Chinese with English abstract)

[29]Kapur J N, Sahoo P K, Wong A K C. A new method for gray-level picture thresholding using the entropy of the histogram[J]. Computer Vision, Graphics, and Image Processing, 1985, 29(3): 273-285.

[30]Woebbecke D M, Meyer G E, Bargen K V, et al. Plant species identification, size, and enumeration using machine vision techniques on near-binary images[J]. SPIE Optics in Agriculture and Forestry, 1992, 1836: 208-219.

[31]Hunt E R, Cavigelli M, Daughtry C S T, et al. Evaluation of digital photography from model aircraft for remote sensing of crop biomass and nitrogen status[J]. Precision Agriculture, 2005, 6(4): 359-378.

[32]Gamon J A, Surfus J S. Assessing leaf pigment content and activity with a reflectometer[J]. New Phytologist, 1999, 143(1): 105-117.

[33]Verrelst J, Schaepman M E, Koetz B, et al. Angular sensitivity analysis of vegetation indices derived from Chris/Proba data[J]. Remote Sensing of Environment, 2008, 112(5): 2341-2353.

[34]Sellaro R, Crepy M, Trupkin S A, et al. Cryptochrome as a sensor of the blue/green ratio of natural radiation in arabidopsis[J]. Plant Physiology, 2010, 154(1): 401-409.

[35]Louhaichi M, Borman M M, Johnson D E. Spatially located platform and aerial photography for documentation of grazing impacts on wheat[J]. Geocarto International, 2001, 16(1): 65-70.

[36]Woebbecke D M, Meyer G E, Von Bargen K, et al. Color indices for weed identification under various soil, residue, and lighting conditions[J]. Transactions of the American Society of Agricultural Engineers (Transactions of the CSAE), 1995, 38(1): 259-269.

[37]Kataoka T, Kaneko T, Okamoto H, et al. Crop growth estimation system using machine vision[C]. Proceedings 2003 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM 2003), Kobe, Japan, 2003, 1079-1083.

[38]Hague T, Tillett N D, Wheeler H. Automated crop and weed monitoring in widely spaced cereals[J]. Precision Agriculture, 2006, 7(1): 21-32.

[39]Meyer G E, Neto J C. Verification of color vegetation indices for automated crop imaging applications[J]. Computers & Electronics in Agriculture, 2008, 63(2): 282-293.

[40]Guijarro M, Pajares G, Riomoros I, et al. Automatic segmentation of relevant textures in agricultural images[J]. Computers & Electronics in Agriculture, 2011, 75(1): 75-83.

[41]Guerrero J M, Pajares G, Montalvo M, et al. Support vector machines for crop/weeds identification in maize fields[J]. Expert Systems with Applications, 2012, 39(12): 11149-11155.

[42]Bendig J, Kang Y, Aasen H, et al. Combining UAV-based plant height from crop surface models, visible, and near infrared vegetation indices for biomass monitoring in barley[J]. International Journal of Applied Earth Observation & Geoinformation, 2015, 39(7): 79-87.

[43]周在明,楊燕明,陳本清. 基于可見(jiàn)光波段無(wú)人機(jī)影像的入侵物種互花米草提取研究[J]. 亞熱帶資源與環(huán)境學(xué)報(bào),2017,12(2):90-95.

Zhou Zaiming, Yang Yanming, Chen Benqing. Study on the extraction of exotic species spartina alterniflora from UAV visible images[J]. Journal of Subtropical Resources and Environment, 2017, 12(2): 90-95. (in Chinese with English abstract)

Vegetation information recognition in visible band based on UAV images

Gao Yonggang1,2,3, Lin Yuehuan1, Wen Xiaole1※, Jian Wenbin1,2, Gong Yingshuang3

(1350116; 2350116,; 3.(),350108,)

Nowadays, UAV (unmanned aerial vehicle) remote sensing has been widely used in various research fields, due to its incomparable advantages over traditional satellite remote sensing, such as lower cost, fast image access, and high spatial resolution, and so on. But most of the vegetation indices are constructed based on visible bands and near-infrared bands of satellite remote sensing images, and few of them are constructed only based on visible-light bands. Thus, it is necessary to construct a universal vegetation index that is suitable for the visible-light bands of UAV images. According to the analysis of the spectral characteristics of 6 kinds of typical features based on regions of interest in visible-light images from UAV images, this paper proposed a new vegetation index based on red, green and blue bands, named Excess green-red-blue difference index (EGRBDI). The formula of EGRBDI was that the sum value between the square of 2 times green band and the product of red and blue bands divided the difference value of them. The value range of EGRBDI was the interval [-1, 1]. To determine the accuracy and reliability of EGRBDI, 18 kinds of vegetation indices had been studied in this paper, such as CIVE GLI, ExG, and so on. The overlap between different object types was obtained by calculating the mean value and 1-fold standard deviation of vegetation indices. The results showed that EGRBDI, GLI, ExG, g, CIVE, RGBVI, and V-MSAVI had no overlap between vegetation and non-vegetation information, while other vegetation indices appeared the different degree of overlap. Moreover, EGRBDI effectively enlarged the identification range of vegetation information and reduced the identification range of non-vegetation information. When the grey histogram of vegetation index existed distinct bimodal peaks, the corresponding discrimination performance of ground features was relatively strong.Therefore, the quantized interval of gray histograms should be normalized to the interval [0, 255] for the comparative analysis between the indices. Results of the analysis concluded that EGRBDI, GLI, ExG, g, CIVE, RGBVI, and V-MSAVI had distinct bimodal-peak characteristics and scarcely appeared thorn peaks in the histogram, but the others had either no obvious bimodal peaks or obvious thorn peaks. To determine the thresholds of vegetation information identification, the bimodal histogram method and the maximum entropy method were used to determine the threshold of each vegetation index and got the optimal threshold of each vegetation index by the precision comparison method. The accuracy evaluation results revealed that GBRI and ExB obtained higher classification accuracy by the maximum entropy method than the bimodal histogram method. WI and VEG had the same accuracy between the two methods, and the other 15 indices did better on the bimodal histogram method. Therefore, the maximum entropy method was used to determine the thresholds of GBRI and ExB, while the other indices used the bimodal histogram method to determine their thresholds in this paper. Through the comparative analysis of the experimental results, it could be found that EGRBDI was generally better than the other 18 algorithms and had a great advantage in the case of the low vegetation coverage, which had a total accuracy of 97.67% and a Kappa coefficient of 0.9415. Another 3 UAV images had been used to extract vegetation information of top 5 higher precision indices to further verify the suitability in the area of various ground subjects and used 400 random points to evaluate the vegetation extraction accuracy. The accuracy of the vegetation and non-vegetation information was not less than 90%. The total accuracy in the 3 study areas was higher than 93%. Additionally, the Kappa coefficient was greater than 0.85. The results showed that EGRBDI had been less affected by ground subjects and shadows, and it had better applicability, reliability, and accuracy of vegetation extraction.

remote sensing; vegetations; spectrum analysis; unmanned aerial vehicle; visible-bands; excess green-red-blue difference index

高永剛,林悅歡,溫小樂(lè),簡(jiǎn)文彬,龔應(yīng)雙. 基于無(wú)人機(jī)影像的可見(jiàn)光波段植被信息識(shí)別[J]. 農(nóng)業(yè)工程學(xué)報(bào),2020,36(3):178-189.doi:10.11975/j.issn.1002-6819.2020.03.022 http://www.tcsae.org

Gao Yonggang, Lin Yuehuan, Wen Xiaole, Jian Wenbin, Gong Yingshuang. Vegetation information recognition in visible band based on UAV images[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(3): 178-189. (in Chinese with English abstract) doi:10.11975/j.issn.1002-6819.2020.03.022 http://www.tcsae.org

2019-12-03

2020-01-18

福建省自然科學(xué)基金(2019J01649);國(guó)家自然科學(xué)基金(41861134011、41801393)

高永剛,副教授,博士,主要從事遙感圖像處理與應(yīng)用、衛(wèi)星大地測(cè)量方面的研究。Email:yggao@fzu.edu.cn

溫小樂(lè),副教授,博士,主要從事遙感圖像處理與應(yīng)用研究。Email:wenxl@fzu.edu.cn

10.11975/j.issn.1002-6819.2020.03.022

P237.3; TP751.1

A

1002-6819(2020)-03-0178-12

猜你喜歡
植被指數(shù)直方圖波段
符合差分隱私的流數(shù)據(jù)統(tǒng)計(jì)直方圖發(fā)布
最佳波段組合的典型地物信息提取
基于無(wú)人機(jī)圖像的草地植被蓋度估算方法比較
冬小麥SPAD值無(wú)人機(jī)可見(jiàn)光和多光譜植被指數(shù)結(jié)合估算
最佳波段選擇的遷西縣土地利用信息提取研究
Bp-MRI灰度直方圖在鑒別移行帶前列腺癌與良性前列腺增生中的應(yīng)用價(jià)值
基于差分隱私的高精度直方圖發(fā)布方法
小型化Ka波段65W脈沖功放模塊
中考頻數(shù)分布直方圖題型展示
植被指數(shù)監(jiān)測(cè)綠洲農(nóng)區(qū)風(fēng)沙災(zāi)害的適宜性分析