Canopy Extraction in a Banana Crop From UAV Captured Multispectral Images

Abstract

In development economics of many countries, banana has been thought of as a key factor, however, banana trees suffer from susceptible to the main climatic conditions such as temperature, humidity, or solar radiation, resulting in a decrease in canopy and leaf area, which has an impact in bunch size and quality. In recent years, teledetection has emerged as a method to analyze the state of the plantation, where multispectral images captured by Unmanned Aerial Vehicle (UAV) are the main information source. This study set out to canopy extraction and estimation of a banana plantation from multispectral images taken at 35 meters height, which are transformed to HSV color field by red-edge band reflectance (REG) and near infrared (NIR). This allowed segmentation to separate the canopy from objects such as dry leaves, ground, and other elements. Finally, the results were compared with manual technique, resulting the proposed methodology has a significant accuracy and adjustment in the results..

Plant phenotyping relevance

UAVマルチスペクトル画像からバナナの樹冠を抽出・推定し、手動法と精度比較しており、植物形態の取得手法が研究の中心である。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.