The Agriculture Innovation using Remote Sensing (AIRS) project intends to combine the areas of artificial intelligence and remote sensing to create a technological solution to monitor the grape leafs in vineyards using satellite images obtained by the European Space Agency. The correct assessment of these variables allows for sustained decisions to be made with an impact on the management of agricultural areas. The implementation of the practices of precision agriculture allows the reduction of the use of pesticide and waste or irrigation water, resulting in a more sustainable agricultural system and the development of rural communities. The innovation of the AIRS project consists in the use of high resolution images acquired by Unmanned Aerial Vehicles to increase the resolution of images coming from the Sentinel-2 satellite, using artificial intelligence. The project foresees the implementation in the vineyards of the members of Adega Cooperativa de Pinhel, with the results later made available to the agricultural community through an online platform.
Climate change has led to a higher incidence of extreme events, which in Portugal has resulted in more storms in winter, hotter summers and longer dry periods. As a result, efficient management of vine cultivation and knowledge about the resilience of vines/varieties has become crucial, with interest in precision agriculture increasing in recent years because it allows accurate monitoring of the entire crop.
Acquiring and maintaining an in situ sensor network is expensive and, therefore, the use of satellite images (e.g. Sentinel-2), available for free, is an asset that should be exploited. Its disadvantage, currently, is the limited spatial resolution, which can vary between 10 and 60 meters, depending on the spectral band that is being used. The objective of this project is to develop super-resolution methods based on recent models in the field of artificial intelligence (e.g., Generative Adversarial Networks) that can be applied to remote sensing data. Furthermore, the project aims to use methods based on deep learning (e.g., Convolutional Neural Networks) to automatically extract phytosanitary metrics from agricultural plantations. Thus, by providing information on – when and where – plants need water, the project will result in an optimisation and efficient management of vineyard irrigation which in turn will lead to greater resistance to variations in weather conditions (more frequent due to climate change). The results will be made available and integrated into a digital services platform for Portugal and provides farmers with an effective way to monitor the health of their crops.
TeroMovigo is leading the project and responsible for performing the observations with the drones over the vineyards. In addition, it will help UBI to develop the super-resolution methodology and prepare the satellite images in collaborations with Food4Sustainability. Finally, it is responsible for the implementation of the web-portal with the results.
University of Beira Interior
Prof. Dr. João Neves of the Department of Informatics at the University of Beira Interior will be responsible for the development of super-resolution of satellite images, using high resolution images taken with drones to train the neural network. In addition, he will supervise the work to identify the plants in the new images and evaluate the indices calculated for these images.
Food4Sustainability has many years of experience with investigating methods to create sustainable food production and will be our main interface with the agriculture community. In the scope of this project, Dr. Cesare Neto, Dr. Ricardo Chagas of F4S will organise regular campaigns to make in-situ measurements that will be used to validate the indices calculated from the satellite images. In addition, Dr. Silvia Moreira of F4S will organise workshops, such as the GROW events, to disseminate the results of the project.
If you have questions about this project, please feel free to contact us: