Integrated system for vineyard yield prediction using drones and spectral imaging
Data
Acronym: VYE SYSTEM
Reference: RTC2019-006949-2
Partners: Instituto Tecnológico y de Energías Renovables (ITER) and Universidad Politécnica de Valencia (UPV).
Duration: 07/01/2020 – 09/30/2022 (27 months)
Budget: 574,725.55 €
Co-Financing: Collaboration Challenges 2019. State Program of R&D&I Oriented to the Challenges of Society.
Project overview
The general objective of VYE SYSTEM is to obtain as a solution an innovative system for predicting the expected yield of a vine crop in a pre-harvest period (at least one month in advance), accurately, at low cost and in real time, which displays its results through a digital platform, allowing producers to optimise the resources involved in this task.
To achieve the described objective, the new VYE SYSTEM system will be designed as a solution, based on the use of unmanned aircraft or ‘drones’, with multispectral cameras incorporated to capture information through images, which will be processed by means of a complex algorithm calculating different parameters and indices linked to the level of crop yields. In order to perform this task, the algorithm will also be composed of neural networks, which will be trained in the field with the calculation of spectral parameters and the contrast of these with the recorded crop yields. Once the data has been processed and the result obtained, a yield map will be generated which can be visualized through the digital platform by means of different devices.
Thus, the new VYE SYSTEM will consist of three distinct subsystems (hardware, communication, and software), along with a digital platform.
- The HARDWARE SUBSYSTEM will consist of a drone as the main element, to which a five-band multispectral camera will be installed. This camera will be responsible for collecting information from the analyzed terrain. An adaptation kit to connect both elements, and some other auxiliary elements, will also be required.
- The COMMUNICATION SUBSYSTEM will be responsible for transferring all the information captured by the hardware subsystem to the cloud, either with an intermediate control center, or directly.
- The SOFTWARE SUBSYSTEM will be responsible for carrying out the processing and analysis of all the received information, in such a way that it results in the prediction of crop yield, discretized by hectare and also in total. It will employ mathematical image processing algorithms as well as Artificial Intelligence (AI) techniques, based on the aforementioned Neural Network.
- The DIGITAL PLATFORM will have the function of displaying results in real time, in a simple and easily understandable manner. It will have different visualization typologies for various devices through which the user can have access, such as computer, mobile phone, etc. In turn, through the digital platform, it will be possible to access historical records of measurements made, data analysis, among others. This will be the main tool of the service offered to end users, that is, a SaaS (Software as a Service) type service will be provided.
Source: The vineyard images have been provided by Tenerife Rural for the VYE SYSTEM project.
Links of interest:
