Identification of weeds from digital images
- Publikations-Art
- Kongressbeitrag
- Autoren
- Weis, M. and Gerhards, R.
- Erscheinungsjahr
- 2007
- Veröffentlicht in
- Agricultural Field Trials - Today and Tomorrow
- Herausgeber
- Bleiholder, Hermann and Piepho, Hans-Peter
- Verlag
- Verlag Grauer , Beuren, Stuttgart
- ISBN / ISSN / eISSN
- 978-3-86186-541-4
- Seite (von - bis)
- 258-263
- Tagungsname
- Landwirtschaftliches Versuchswesen -- heute und morgen
- Tagungsort
- Stuttgart-Hohenheim
- Tagungsdatum
- 08.-10. October 2007
Weed control is one of the classical treatments in many kinds of agricultural
production systems. Site-specific weed control based on weed maps has raised
interest in the precision farming community in the last years. Weeds are spatially
inhomogeneous in their distribution in the field (Gerhards et al., 1997, Mortensen
et al., 1998, Gerhards & Christensen, 2003). High weed densities in patches
alternate with lower or no weed density in other areas. The patches remain stable
over years due to high local seed densities in seed banks (Wilson & Brain, 1991).
Management zones can be set up based on information about the weed distribution.
A high weed density has a negative influence on the yield of the cultivated crop.
Weed can be controlled with herbicide applications. The estimation of the weed
distribution is a crucial step for the creation of weed density maps and the site-
specific management. Site-specific herbicide application can be used to manage
weeds according to their abundance and variability.