Simafield: a software to spatially simulate agronomic images

Abstract : Environmental measures lead farmers to use more and more decision support tools to reduce their chemical inputs without losing yield. Imaging system based on crop/weed discrimination algorithms have been developed to detect weeds in field and allow a site specific spraying. These algorithms needs to be thoroughly evaluated before considering a public release and current evaluation are based on real images and are very tedious. SimAField is a software that aims at modeling agronomic fields with crop and weeds and simulate the picture taking process to replace ground truth and allow an automated evaluation of crop/weed discrimination algorithms.
Complete list of metadatas

https://hal-agrosup-dijon.archives-ouvertes.fr/hal-01744270
Contributor : Administrateur Agrosupdijon <>
Submitted on : Tuesday, March 27, 2018 - 11:48:07 AM
Last modification on : Tuesday, January 15, 2019 - 2:44:02 PM

Identifiers

  • HAL Id : hal-01744270, version 1

Citation

Gawain Jones, Christelle Gée, Sylvain Villette, Jean-Noël Paoli. Simafield: a software to spatially simulate agronomic images. 22. Conférence du Columa, Journées internationales sur la lutte contre les mauvaises herbes, Dec 2013, Dijon, France. ⟨hal-01744270⟩

Share

Metrics

Record views

84