Weed discrimination using ultrasonic sensors

Autor: Andújar, Dionisio, Escolà i Agustí, Alexandre, Dorado, José, Fernández-Quintanilla, César
Přispěvatelé: Comisión Interministerial de Ciencia y Tecnología, CICYT (España), Andújar, Dionisio, Escolà i Agustí, Alexandre, Dorado, José, Fernández-Quintanilla, César, Andújar, Dionisio [0000-0002-5801-0944], Escolà i Agustí, Alexandre [0000-0002-9775-5471], Dorado, José [0000-0002-2268-2562], Fernández-Quintanilla, César [0000-0002-2886-9176]
Rok vydání: 2011
Předmět:
Zdroj: Repositorio Abierto de la UdL
Universitad de Lleida
Recercat. Dipósit de la Recerca de Catalunya
instname
Digital.CSIC: Repositorio Institucional del CSIC
Consejo Superior de Investigaciones Científicas (CSIC)
Digital.CSIC. Repositorio Institucional del CSIC
DOI: 10.1111/j.1365-3180.2011.00876.x
Popis: A new approach is described for automatic discrimination between grasses and broad-leaved weeds, based on their heights. An ultrasonic sensor was mounted on the front of a tractor, pointing vertically down in the inter-row area, with a control system georeferencing and registering the echoes reflected by the ground or by the various leaf layers. Static measurements were taken at locations with different densities of grasses (Sorghum halepense) and broad-leaved weeds (Xanthium strumarium and Datura spp.). The sensor readings permitted the discrimination of pure stands of grasses (up to 81% success) and pure stands of broad-leaved weeds (up to 99% success). Moreover, canonical discriminant analysis revealed that the ultrasonic data could separate three groups of assemblages: pure stands of broad-leaved weeds (lower height), pure stands of grasses (higher height) and mixed stands of broad-leaved and grass weeds (medium height). Dynamic measurements confirmed the potential of this system to detect weed infestations. This technique offers significant promise for the development of real-time spatially selective weed control techniques, either as the sole weed detection system or in combination with other detection tools.
This research was funded by the Spanish CICyT (project AGL 2008-04670-C03).
Databáze: OpenAIRE
Externí odkaz: