English   español  
Please use this identifier to cite or link to this item: http://hdl.handle.net/10261/164094
Share/Impact:
Statistics
logo share SHARE   Add this article to your Mendeley library MendeleyBASE
Visualizar otros formatos: MARC | Dublin Core | RDF | ORE | MODS | METS | DIDL
Exportar a otros formatos:

Title

Generation of canopy height model based on point clouds and spectral data: a case study on grapevine

AuthorsMesas-Carrascosa, Francisco Javier; Peña Barragán, José Manuel ; Castro, Ana Isabel de ; Torres-Sánchez, Jorge ; Jiménez-Brenes, Francisco Manuel ; García-Ferrer, Alfonso; López Granados, Francisca
KeywordsDigital elevation mode
Classification cloud points
Canopy height model
Index color
Issue DateJun-2017
Citation5th International Conference on Small Unmanned Aerial Systems for Environmental Research (2017)
AbstractThe availability of very high density cloud points is of increasing interest for scientists and oth er users involved in obtaining precise information for environmental, forestry or agronomical processes, among others. In the context of precision viticulture, UAV images are a potential way to map crop structure parameters, such as height row or vegetation cover fraction. To derive the structural information a very dense point cloud is extracted from UAV images (structure from motion). Every point contains X, Y and Z coordinates and spectral values of Red (R), Green (G) and Blue (B) bands from images can be assigned to the cloud points. Once cloud points are derived, they need to be processed and filtered to extract information of interest. Over the past years different filtering techniques have been developed to classify points from these cloud points. First, bare earth points are classified and a Digital Elevation Model (DEM) is generated. Then, the remaining points are classified as, for example, low, medium or high vegetation or building s, producing a Digital Surface Model (DSM). The difference between a DSM and DEM yields a Canopy Height Model (CHM). Therefore, it is necessary to distinguish terrain points from the rest of typologies. To classify points as terrain most filtering techniques assume that the Earth’s surface is continuous in all directions and apply geometric or morphological constraints. The methodology herein presented is based on the use of RGB spectral information associated with every point to discriminate between vegetation and other classes by colour indexes without using any geometric conditions. To validate the proposed methodology, cloud points from images taken with an RGB camera on-board an unmanned aerial vehicle (UAV) have been used. The UAV flights were performed at a vineyard whose land uses were grapevine rows, bare soil and cover crops (mainly composed by grass species) between the rows. The cloud points linked to this spectral information were accurately classified and a DEM, DSM and CHM were generated. These results are very useful to describe the structure of vineyards.
DescriptionTrabajo presentado en la 5th International Conference on Small Unmanned Aerial Systems for Environmental Research (UAS4Enviro2017), celebrada en Vila Real (Portugal) del 28 al 30 de junio de 2017.
URIhttp://hdl.handle.net/10261/164094
Appears in Collections:(IAS) Comunicaciones congresos
Files in This Item:
File Description SizeFormat 
accesoRestringido.pdf15,38 kBAdobe PDFThumbnail
View/Open
Show full item record
Review this work
 


WARNING: Items in Digital.CSIC are protected by copyright, with all rights reserved, unless otherwise indicated.