Cellesprøjtning af ukrudt i majs
Summary and conclusions
A sensor-based spraying system with an integrated unit for automatic sampling of technical spraying data was constructed and analysed in the project ”Cell spraying of weeds in maize”. This spraying system was named cell spraying because it sprays solely in areas (cells) in the field in which there are weeds. All other areas (cells), that do not contain weeds, are not sprayed.
Since the occurrence of weeds in the field varies significantly and weeds often grow in patches, using this spraying system means that loss on the soil surface will be greatly reduced.
1.5 Background and aims
The background for this project is the challenge of meeting the goals of the National Pesticide Action Plan 2004-2009. Optimisation of herbicide choice, dosage, mechanical methods and conventional spraying techniques do not seem to be able to offset the general increase in treatment frequency.
The aim of the project is to reduce herbicide use in maize by 50 percent by spraying only in areas with weeds. The goal was to investigate how spraying can be adapted to the variations in weed occurrence in the field by use of vision-based spraying systems.
1.6 The studies
The project was organised in five work packages that together created a whole including system construction, optimisation and documentation.
The work involving analysis, development and integration of the vision and spraying system was carried out at the Department of Agricultural Engineering at the University of Aarhus, while the biological efficacy studies were carried out at the University’s Department of Integrated Pest Management.
The vision-based spraying system was expanded to include automatic documentation of the plant protection that was carried out, which was carried out by the University of Aarhus and the Danish Agricultural Advisory Services.
1.7 Main conclusions
The project has shown that vision-based spraying technique can treat the surface of the field in small cells with a dosage requirement that does not differ significantly from the dosage used for conventional boom spraying.
Theoretical analyses have shown that spraying cells measuring 0.01 m² can potentially reduce pesticide use by 50-70 percent compared to conventional boom spraying.
Under good light conditions the cell sprayer can operate with a forward speed of approximately 1 m/s corresponding to 3.6 km/hour.
The project has shown that it is possible to build up a data sampling system that can document plant protection actions in real time under practical conditions using automatic data collection and transfer to a central server.
1.8 Project results
The cell sprayer has a set of video cameras in a vision system which takes images of the soil surface immediately in front of the spraying boom. The images are analysed for the occurrence of weeds. When one or more weeds are found in the image, the information about their location is saved.
The image is divided into rectangular units (cells) of 30 mm in the driving direction and 107 mm at a right angle to the driving direction. Since the cameras are fixed in relation to the spraying boom, the cells are placed so that the nozzles – with a certain time lapse – pass over the middle of each cell.
The cameras are placed at a height above the soil so that the images include six cell strips in the driving direction and where each strip contains seven cells at right angles to the driving direction. Each image is thus divided in 6x7 = 42 cells, which are analysed individually.
If weeds are detected, then the cell in question is marked for later spraying. Otherwise, the cell is not sprayed. On the basis of this information a small spraying map is prepared (Figure 4).
Figure 4. Schematic illustration of the cell spraying system for one camera
How far it is possible to spray against weeds efficiently in small cells was investigated using cell spraying of camomile with Roundup Bio in three concentrations (n=0,45 l/ha).
The spraying was carried out in cells measuring 10 x 10 cm and with a driving speed of 1 m/s. The results from the innermost 8.1 x 8.1 cm are shown in Figure 5. The spraying nozzle TeeJet 6501-E was also tested at 2 m/s. The values found correspond to effects of 95-98 percent at the highest dosage and 75-95 percent at the lowest dosage. There were minor differences between the nozzles and the effect at 2 m/s was on equivalent with the effect at 1 m/s.
Figure 5. Fresh weight of camomile using cell spraying with Roundup Bio in three dosages (n=0,45 l/ha). The results are from the innermost 8.1 x 8.1 cm. Untreated on same area = 2.36 g (length of Y-axis). LSDdos1,0 = 0.041. LSDdos0,6 = 0.089. LSDdos0,36 = 0.26.
Various scenarios were used in connection with the theoretical calculations of the potential for cell spraying in maize.
The actual potential for the method must be validated in field trials but on the basis of the described results it can be concluded that cell spraying in cells measuring 0.01 m² has a potential for a pesticide use reduction of at least 50-70 percent based on the given conditions.
The spraying system was actually developed with a cell size of 30 x 107 mm as a starting point, which increases the potential for reduction even more. With a weed density of 220 plants per m² occurring in patches and a soil effect (residual effect) included in the calculations, the above-mentioned theoretical calculations result in a potential reduction of 80–90 percent.
Version 1.0 Oktober 2008, © Miljøstyrelsen.