P11: Digital and technology IV
E-poster session | Chairperson: Joaquin Cochero | Tuesday, 08 September 2020 | 18:30-18:45
18:30-18:35 | P11.1 Combining citizen science and machine learning techniques to address the two main issues of user engagement and data validation in citizen science projects
Author/s: Maryam Lotfian and Jens Ingensand
Presenter: Maryam LotfianSummary:
Nowadays there are many citizen science applications to collect biodiversity observations, and this has resulted in freely accessible large species datasets. However, it is essential to understand how to validate such large datasets taking into account species characteristics.We have developed a citizen science application to collect biodiversity observations, which includes a real-time automatic filtering functionality to flag rare or erroneous observations with the aid of machine learning techniques.The objective of this automatic filtering and real-time feedback is on the one hand to improve data quality, and on the other, to sustain participation by giving informative feedback to the volunteers.
18:35-18:40 | P11.2 Crowdsourced and surveyed images as a source of in-situ data for crops, phenology, and agricultural landscapes
Author/s: Marijn van der Velde, Raphaël d'Andrimont, Momchil Iordanov, Laura Sanchez Martinez and Guido Lemoine
Presenters: Marijn van der Velde, Laura Martinez SanchezSummary:
Use of crowd and expert surveyed imagery in combination with machine learning algorithms to extract geo-located information on crops and phenology. High-quality in-situ data is needed to complement and extract information from the near-instantaneous and time-dense Earth Observations (EO) by the Copernicus Sentinel 1 and 2.
For this purpose, experts collected SLI using side-looking action cameras for a fixed route survey in Flevoland (NL) during eight dates in the 2018 growing season. This resulted in a SLI dataset of nearly 200,000 full HD resolution images. At more than 200 parcels detailed phenological observations were made for a range of crops. This information provides training data to extract crop type and phenology from the SLI images with machine learning algorithms.
18:40-18:45 | P11.3 Building a European community of interest around invasive alien species
Author/s: Eugenio Gervasini, Celia López Cañizares and Ana Cristina Cardoso
Presenter: Eugenio GervasiniSummary:
Invasive Alien Species (IAS) affect all environments including cities. Citizens can greatly contribute to their detection, monitoring and management, supporting the implementation of the European policy on biodiversity. The European Alien Species Information Network (EASIN), disseminates updated scientific information on alien species in Europe, and is committed to Citizen Science, via a dedicated App, a communication strategy, including Social Media, to create a European-wide community of interest on IAS. Interactions with the public allow a better and personalized feedback, help improving communication, update the App features, and promotes people’s long term engagement in early detection, monitoring, and management of IAS.