利用無人機影像構(gòu)建作物表面模型估測甘蔗LAI
發(fā)布時間:2018-05-17 16:40
本文選題:遙感 + 無人機。 參考:《農(nóng)業(yè)工程學(xué)報》2017年08期
【摘要】:為探討從作物表面模型(crop surface models,CSMs)中提取株高來估算糖料蔗葉面積指數(shù)(leaf area index,LAI)的可行性,該文采用無人機-RGB高清數(shù)碼相機構(gòu)成的低空遙感平臺,以廣西糖料蔗為研究對象,采集了糖料蔗全生育期的高清數(shù)碼影像,分別在有無地面控制點條件下建立各生育期CSMs并提取株高。此外,該文利用高清數(shù)碼影像計算了6種可見光植被指數(shù)并建立LAI估算模型,用以對比從CSMs提取的株高對LAI的估算效果。結(jié)果表明:全生育期CSMs提取的株高與實測株高顯著相關(guān)(P0.01),株高預(yù)測值與實測值高度擬合(R2=0.961 2,RMSE=0.215 2)。選取的6種可見光植被指數(shù)中,綠紅植被指數(shù)對糖料蔗伸長末期以前的LAI的估測效果最好(R2=0.779 0,RMSE=0.556 1,MRE=0.168 0)。相同條件下,株高對LAI有更高的估測精度,其中CSMs提取的株高估測效果優(yōu)于地面實測株高,預(yù)測模型R2=0.904 4,RMSE=0.366 2,MRE=0.124 3。研究表明,使用無人機拍攝RGB影像來提取株高并運用于糖料蔗重要生育期LAI的估算是可行的,CSMs提取的株高擁有較高的精度。該研究可為大區(qū)域進行精準快速的農(nóng)情監(jiān)測提供參考。
[Abstract]:In order to study the feasibility of estimating leaf area index (Lai) from crop surface model surface models CSMs, a low altitude remote sensing platform based on UAV RGB high-definition digital camera was used to study sugarcane in Guangxi. The high-definition digital images of sugarcane growth period were collected, and the CSMs of each growth period was established under the condition of ground control point or not and the plant height was extracted. In addition, six kinds of visible light vegetation indices were calculated by using high-definition digital images and LAI estimation models were established to compare the effect of plant height extracted from CSMs on LAI estimation. The results showed that the plant height extracted by CSMs at the whole growth stage was significantly correlated with the measured plant height (P0.01). Of the 6 visible light vegetation indices selected, the green red vegetation index had the best effect on estimating LAI of sugarcane before the end of sugarcane elongation. Under the same conditions, the plant height had higher estimation accuracy to LAI, and the plant height estimated by CSMs extraction was better than that of ground measured plant height, and the prediction model R2O0.9044 RMSEN 0.366 2 MREE 0.124 3 124. The results show that it is feasible to extract plant height from RGB images and estimate LAI in sugar cane at important growth stage. This study can provide a reference for accurate and rapid monitoring of agricultural conditions in large areas.
【作者單位】: 武漢大學(xué)水資源與水電工程科學(xué)國家重點實驗室;廣西壯族自治區(qū)水利科學(xué)研究院;
【基金】:高等學(xué)校全國優(yōu)秀博士學(xué)位論文作者專項資金(201248) 廣西水利廳科技項目(201615)
【分類號】:S127;S566.1
,
本文編號:1902094
本文鏈接:http://sikaile.net/kejilunwen/nykj/1902094.html
最近更新
教材專著