Addressing the intrinsic variability within vineyards is a key factor to perform precision viticulture management. To this aim, new and more reliable methods for vineyard monitoring purposes must be defined. The introduction of Unmanned Aerial Vehicle (UAV) airborne sensors makes available a considerable amount of data with very high resolution, in terms of both spatial and temporal dimension. In this work, a data fusion approach for vigour characterization in vineyards is presented, which exploits the information provided by 2D multispectral aerial imagery, 3D point cloud crop models and aerial thermal imagery. A crucial phase of the procedure is the proper management of data provided by several sources, to achieve high consistency of the obtained huge dataset. The enhanced effectiveness of the proposed method to classify vines in different vigour classes exploiting multi source data was proved by an experimental campaign, considering 30 portions of vine rows, each made by 8 vines. Results showed that the error of the discriminant analysis using data fusion reach an improvement ranging from 67% to 90% with respect to a single data source, with a misclassification error rate of 3%.

2D and 3D data fusion for crop monitoring in precision agriculture

Lorenzo Comba
First
;
Alessandro Biglia;Davide Ricauda Aimonino;Paolo Barge;Cristina Tortia;Paolo Gay
Last
2019

Abstract

Addressing the intrinsic variability within vineyards is a key factor to perform precision viticulture management. To this aim, new and more reliable methods for vineyard monitoring purposes must be defined. The introduction of Unmanned Aerial Vehicle (UAV) airborne sensors makes available a considerable amount of data with very high resolution, in terms of both spatial and temporal dimension. In this work, a data fusion approach for vigour characterization in vineyards is presented, which exploits the information provided by 2D multispectral aerial imagery, 3D point cloud crop models and aerial thermal imagery. A crucial phase of the procedure is the proper management of data provided by several sources, to achieve high consistency of the obtained huge dataset. The enhanced effectiveness of the proposed method to classify vines in different vigour classes exploiting multi source data was proved by an experimental campaign, considering 30 portions of vine rows, each made by 8 vines. Results showed that the error of the discriminant analysis using data fusion reach an improvement ranging from 67% to 90% with respect to a single data source, with a misclassification error rate of 3%.
2019 IEEE International Workshop on Metrology for Agriculture and Forestry (MetroAgriFor)
Institute of Electrical and Electronics Engineers Inc.
62
67
9781728136110
http://ieeexplore.ieee.org/xpl/mostRecentIssue.jsp?punumber=8902128
Precision viticulture, Crop modelling, 2D and 3D data fusion, UAV remote sensing
Lorenzo Comba, Alessandro Biglia, Davide Ricauda Aimonino, Paolo Barge, Cristina Tortia, Paolo Gay
File in questo prodotto:
File Dimensione Formato  
08909219.pdf

Accesso riservato

Descrizione: pdf editoriale
Tipo di file: PDF EDITORIALE
Dimensione 419.46 kB
Formato Adobe PDF
419.46 kB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/2318/1720327
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 17
  • ???jsp.display-item.citation.isi??? ND
social impact