Pl@ntNet Crops: merging citizen science observations and structured survey data to improve crop recognition for agri-food-environment applications. (1st February 2023)

• Record Type: Journal Article
• Authors: van der Velde, M; Goeau, H; Bonnet, P; d'Andrimont, R; Yordanov, M; Affouard, A; Claverie, M; Czucz, B; Elvekjaer, N; Martinez-Sanchez, L; Rotllan-Puig, X; Sima, A; Verhegghen, A; Joly, A
• Is Part Of: Environmental research letters. Volume 18:Number 2(2023)
• Publication Date: 2023-02-01
• Subjects: citizen science -- deep learning -- computer vision -- authoritative data -- crowdsourcing -- common agricultural policy -- monitoring; Environmental sciences -- Periodicals; Human ecology -- Research -- Periodicals; Environmental health -- Periodicals; 333.7
• DOI: 10.1088/1748-9326/acadf3
• Languages: English
• ISSNs: 1748-9326

Abstract

Abstract: We present a new application to recognize 218 species of cultivated crops on geo-tagged photos, 'Pl@ntNet Crops'. The application and underlying algorithms are developed using more than 750k photos voluntarily collected by Pl@ntNet users. The app is then enriched by data and photos coming from the European Union's (EU) Land Use and Coverage Area frame Survey (LUCAS). During five tri-annual LUCAS campaigns from 2006 to 2018, 242 476 close-up 'cover' photos of crops were collected. The survey protocol for these photos specified that 'the picture should be taken at a close distance, so that the structure of leaves can be clearly seen, as well as flowers or fruits'. This unique labelled data provides an opportunity to further generalize the Pl@ntNet computer vision algorithms to recognize crops and enlarge their geographic representivity across the EU. To include LUCAS cover photos, we semantically match Pl@ntNet species and LUCAS legends, predict the species on LUCAS cover photos with the existing Pl@ntNet algorithm, and consider the accuracy of the classification and the number of species enriched by the photos. By setting a threshold of > 0.5 on the Pl@ntNet prediction probabilities, 70 170 LUCAS photos representing 101 species classified with an accuracy of 0.9 were added to the 'Crops' app. The thematic accuracy of the legacy LUCAS data was improved by distinguishing 218 species, opposed to the original 36 LUCAS levels. Official and publicly financed LUCAS datastreams can now be improved because of Pl@ntNet citizen science, photo collection, and deep learning model development. Further use of the app and policy-relevant workflows in the agri-food-environment domain are discussed.