Articles
REMOTE SENSING IN MONITORING THE DIEBACK OF HAZELNUT ON THE ‘MONTI CIMINI’ DISTRICT (CENTRAL ITALY)
Article number
845_80
Pages
521 – 526
Language
English
Abstract
During recent years, new methodologies for remote analysis, such as remote and proximal sensing, have been added to the various standard techniques for plant disease monitoring.
In particular, multi-spectral images, acquired at a high radiometric and geometric resolution, by means of digital airborne systems or by satellite, can be useful for monitoring some diseases of woody plants.
This work reports the initial results of the application of one of these systems, called A.Sp.I.S. (Advanced SPectroscopic Imaging System) for monitoring the Dieback of hazelnut, in the province of Viterbo, as a case study.
This is a bacterial disease which has affected hazelnut orchards in the province of Viterbo since the 1980s.
The spatial data sources were aerial photographs taken during a decade of growing seasons, from 1996 to 2006. Moreover, satellite images (Quick Bird) were used in order to previously scan wide areas of the territory affected by the disease.
The satellite images, once overlaid and manipulated by means of dedicated software (Image calculator IDRISI), were used as a basis for drawing new elaborated differential digital images.
The digital elaboration showed annually the plants which were suspected of dying the season before.
The suspected plants were then automatically identified by means of E-cognition, a segmentation GIS software which performs object-oriented image analysis.
The results were then confirmed, comparing the A.Sp.I.S. images and by proper ground activities performed in loco.
The software was able to recognize the main part of dead or wilting plants when applied to the images.
Finally, a protocol of investigation was set up, by means of vectorial GIS, in order to follow in time and in space the local spread of the disease.
In particular, multi-spectral images, acquired at a high radiometric and geometric resolution, by means of digital airborne systems or by satellite, can be useful for monitoring some diseases of woody plants.
This work reports the initial results of the application of one of these systems, called A.Sp.I.S. (Advanced SPectroscopic Imaging System) for monitoring the Dieback of hazelnut, in the province of Viterbo, as a case study.
This is a bacterial disease which has affected hazelnut orchards in the province of Viterbo since the 1980s.
The spatial data sources were aerial photographs taken during a decade of growing seasons, from 1996 to 2006. Moreover, satellite images (Quick Bird) were used in order to previously scan wide areas of the territory affected by the disease.
The satellite images, once overlaid and manipulated by means of dedicated software (Image calculator IDRISI), were used as a basis for drawing new elaborated differential digital images.
The digital elaboration showed annually the plants which were suspected of dying the season before.
The suspected plants were then automatically identified by means of E-cognition, a segmentation GIS software which performs object-oriented image analysis.
The results were then confirmed, comparing the A.Sp.I.S. images and by proper ground activities performed in loco.
The software was able to recognize the main part of dead or wilting plants when applied to the images.
Finally, a protocol of investigation was set up, by means of vectorial GIS, in order to follow in time and in space the local spread of the disease.
Publication
Authors
A. Fabi, L. Varvaro
Keywords
Corylus avellana, dieback, Geographic Information System, hazelnut diseases, phytopathogenic bacteria, remote sensing
Online Articles (125)