Monitoring of peanut leaves chlorophyll content based on drone-based multispectral image feature extraction

Abstract

Peanuts are an important cash crop in the national economy, and their chlorophyll content can reflect the fertilizer status of peanuts. Therefore, it is urgent to obtain a rapid and accurate method for monitoring chlorophyll content. In this study, two types of peanuts, Yanghua 1 and Yueyou 45, were planted according to different plant densities, and eight vegetation indices were calculated using Parrot Bluegrass multispectral drone photography, so as to establish different monitoring model, and to compare the accuracy of each model according to the coefficient of determination, root mean square error, and absolute mean error. The chlorophyll content of Yanghua 1 peanut leaves was generally higher than that of Yueyou 45, and the plant density had a greater influence on Yanghua 1 peanut. The one-dimensional linear regression models of NDVI(Normalized Difference Vegetation Index) and GNDVI(Green Normalized Difference Vegetation Index) had a much higher degree of fit and precision than the other indices; and the multiple linear regression model had a more accurate predictive ability than the one-dimensional linear regression model. After testing, BP neural network(Back Propagation Neural Network)is the most suitable model for monitoring the chlorophyll content of peanut with better fit and accuracy than the random forest model. The multispectral drone could rely on the highly accurate predictive model to quickly obtain information on the chlorophyll content in the field and infer the most suitable crop type and planting density for local planting.

Plant phenotyping relevance

ドローン multispectral 画像から植葉のクロロフィル含量を推定する特徴抽出・回帰モデルを開発し、精度比較と検証を行っており、フェノタイピング手法が中心である。

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.