• Title of article

    Design and application of a weed damage model for laser-based weed control

  • Author/Authors

    Christian Marx، نويسنده , , Stephan Barcikowski، نويسنده , , Michael Hustedt، نويسنده , , Heinz Haferkamp، نويسنده , , Thomas Rath، نويسنده ,

  • Issue Information
    روزنامه با شماره پیاپی سال 2012
  • Pages
    10
  • From page
    148
  • To page
    157
  • Abstract
    Horticultural weed control strategies based on chemical and thermal methods are environmentally and energetically questionable. A promising alternative appears to be the use of laser technology. This study evaluates the influence of CO2 laser radiation (10,600 nm) taking into account three laser spot diameters, three laser spot positions and six laser intensities on three growth stages of two weed species (monocotyledonous: Echinochloa crus-galli, dicotyledonous: Amaranthus retroflexus). The lethal impact of irradiation was characterised by a decrease of the weed fresh mass of 90% compared to untreated plants two weeks after irradiation. Weed-specific laser damage models were developed and validated, mapping the probabilities of success (psuccess) of the laser application. Selective lethal laser doses with psuccess = 0.95 were determined. The results showed that lethality was greatest if treatment was performed at early growth stages with high intensity. Monocotyledonous 2-leaf-plants were damaged at high energy levels, whilst 4-leaf-plants were difficult to kill. Dicotyledonous 2-leaf-plants were already damaged at moderate intensities. Thus, the damage of monocots required higher minimum laser doses than the damage of dicots. The influence of the spot position was important, as the unfocused treatment resulted in a decrease in lethality. In combination with robotics and image processing, the damage models developed here can be used to develop laser-based weed control. A model simulation of two driving concepts for the laser system showed that accurate aiming at the meristem as well as specific parameter adjustment is required for an efficient weed control.
  • Journal title
    Biosystems Engineering
  • Serial Year
    2012
  • Journal title
    Biosystems Engineering
  • Record number

    1267829