You are in:

 
 

  USER'SAREA
user ID
password
Forgot your password?
Forgot your username?

 

cabecera noticias img EN


Thursday, 24 May 2018 09:52

Discover an algorithm that allows to obtain strains vineyards information from drone images

Researchers of the imaPing group 'Remote sensing applied to precision agriculture', attached to the ceiA3 and integrated in the Sustainable Agriculture Institute (IAS-CSIC) of Córdoba, have developed an algorithm that allows the automatic obtaining of detailed and individualized information of vineyard strains through the analysis of obtained images with a drone or UAV.
As explained by the researcher of the imaPing group 'Remote sensing applied to precision agriculture', Ana Isabel de Castro, "this achievement is an improvement that will allow the farmer to carry out localized management strategies and an increase of environmental and economic benefits, accordingly".

Traditionally, the strains canopy or other woody crops dimensions have been obtained through manual work becoming an arduous task along intensive working days in the field.

To carry this study out, a dron or UAV, were used to obtain the images. It was attached a low-cost camera to the unmanned aerial vehicle and many pictures were taken flying over three commercial vineyards in Lérida, owned by the Raimat company.

The obtained images were incorporated into a software that automatically carried out the generation of the three-dimensional model of the crop. Finally, an automatic algorithm for object-based image analysis (OBIA) was developed, which allowed us to obtain concrete data on vine strains related to projection of the strain canopy, height and volume from the 3D model obtained.

Algorithm validation

To validate the algorithm generated by the researchers, georeferenced points were established in each of the fields analysed. The height of the vine strains was measured manually on these georeferenced points. Besides, over these points, it was established some squares of 2 x 2 meters where a manual digitalization of the vineyard was carried out thanks to the high spatial resolution of the UAV obtained images.

These values were compared with the ones provided by the algorithm in these areas and related to the vineyard height and classification values, reaching a high degree of precision.
The research was carried out on three vineyards in two different phases: the first in July, when the canopy of the vines was fully developed, and the second in September, after harvesting, when the canopy was less dense. According to the co-author of the work words, Jorge Torres Sánchez, this approach "allowed analysing a wide variability and the results obtained demonstrated the robustness of the algorithm".

Concerning the precise time of execution and processing of the main tasks of the investigation, the UAV flights were developed in interval time of 6 -7 minutes per analysed hectare. On the other hand, the software needed to carry out the 3D automatic reconstruction of the crop, took almost 3 hours (this may vary depending on the capacity of the computer equipment used).
Finally, the analysis of these images was implemented through OBIA automatic algorithm designed in this research, classifying with high precision each of the strains, which had a duration of around 15 minutes per hectare of vineyard analysed.

This study is supported by previous research executed by the imaPing group, led by the researcher Francisca López Granados, in which UAV technology and OBIA analysis have been used for woody crops 3D characterization.

The research is funded within the project framework AGL2017-83325-C4-4R of the Ministry of Economy, Industry and Competitiveness, FEDER Funds). Ana Isabel de Castro and José Manuel Peña research were financed by "Juan de La Cierva" and "Ramón y Cajal" Programs, respectively.

Reference:

Ana I. de Castro, Francisco M. Jiménez-Brenes, Jorge Torres-Sánchez, José M. Peña, Irene Borra-Serrano, Francisca López-Granados. '3-D Characterization of Vineyards Using a Novel UAV Imagery-Based OBIA Procedure for Precision Viticulture Applications' in Remote Sensing. https://doi.org/10.3390/rs10040584

Read 3829 times
  ceiA3   research & transfer
  academic programs   internationalization   what's new on ceiA3
  what is ceiA3?
who takes part in?
governance
contact us
agreements
what do we research?
research groups
website for the researcher
clusters
talented people of the ceiA3
partner enterprises
international doctorate school
grade
masters
training network courses
international mobility
international partners
activities
conferences
blogs of Science
summer scientific campus

Project funded by the Spanish Ministry of Science, Innovation and Universities / Spanish Ministry of Economy and Business Support / Junta de Andalucía within the Campus of International Excellence Framework Program
Legal Notice | Follow us: logo_linkedin
Icono  de conformidad con el Nivel Doble-A,      de las Directrices de  Accesibilidad para el      Contenido Web 1.0 del W3C-WAI
Universidad de Almería Universidad de Cádiz Universidad de Córdoba Universidad de Huelva Universidad de Jaén CSIC Junta de Andalucía - Consejería de Agricultura y Pesca Ministerio de Economía y Empresa Ministerio de Ciencia, Innovación y Universidades Junta de Andalucía Unión Europea Santander Universidades