RoboWeedSupport-Sub Millimeter Weed Image Acquisition in Cereal Crops with Speeds up till 50 Km/H

For the past three years, the Danish project,
RoboWeedSupport, has sought to bridge the gap between the potential
herbicide savings using a decision support system and the required
weed inspections. In order to automate the weed inspections it is
desired to generate a map of the weed species present within the
field, to generate the map images must be captured with samples
covering the field. This paper investigates the economical cost of
performing this data collection based on a camera system mounted
on a all-terain vehicle (ATV) able to drive and collect data at up to 50
km/h while still maintaining a image quality sufficient for identifying
newly emerged grass weeds. The economical estimates are based on
approximately 100 hectares recorded at three different locations in
Denmark. With an average image density of 99 images per hectare
the ATV had an capacity of 28 ha per hour, which is estimated to cost
6.6 EUR/ha. Alternatively relying on a boom solution for an existing
tracktor it was estimated that a cost of 2.4 EUR/ha is obtainable under
equal conditions.

Authors:

• Morten Stigaard Laursen
• Rasmus Nyholm Jørgensen
• Mads Dyrmann
• Robert Poulsen

Keywords:

• Weed mapping
• integrated weed management
• weed recognition.

References:

[1] S. Haneklaus, H. Lilienthal, and E. Schnug, “25 years precision
agriculture in germany–a retrospective,” in 13th International
Conference on Precision Agriculture.
[2] S. Christensen, H. T. Søgaard, P. Kudsk, M. Nørremark, I. Lund, E. S.
Nadimi, and R. Jørgensen, “Site-specific weed control technologies,”
Weed Res., vol. 49, no. 3, pp. 233–241, 1 Jun. 2009.
[3] R. Gerhards, “Site-Specific weed control,” in Precision in Crop Farming.
Springer Netherlands, 2013, pp. 273–294.
[4] C. Gutjahr, M. S¨okefeld, and R. Gerhards, “Evaluation of two patch
spraying systems in winter wheat and maize,” Weed Res., vol. 52, no. 6,
pp. 510–519, 1 Dec. 2012.
[5] D. L. Shaner and H. J. Beckie, “The future for weed control and
technology,” Pest Manag. Sci., vol. 70, no. 9, pp. 1329–1339, Sep. 2014.
[6] R. B. Brown and S. D. Noble, “Site-specific weed management: sensing
requirements— what do we need to see?” Weed Sci., vol. 53, no. 2, pp.
252–258, 2005.
[7] M. Weis, C. Gutjahr, V. R. Ayala, R. Gerhards, C. Ritter, and
F. Sch¨olderle, “Precision farming for weed management: techniques,”
Gesunde Pflanzen, vol. 60, no. 4, pp. 171–181, 2008.
[8] S. Trengove, “Weed mapping,” Developments & Demos, vol. 12, no. 2,
pp. 20–22, 2016.
[9] H. Leithold, “Digital farming I using WEED SENSORS for efficient
weed control,” 7 Mar. 2016.
[10] “AmaSpot sensor nozzle system,” http://press.lectura.de/en/amaspot-sensor-nozzle-system/22092,
accessed: 2015-10-2.
[11] M. Laursen, R. Jørgensen, H. Midtiby, K. Jensen, M. Christiansen,
T. Giselsson, A. Mortensen, and P. Jensen, “Dicotyledon weed
quantification algorithm for selective herbicide application in maize
crops,” Sensors, vol. 16, no. 11, p. 1848, 4 Nov. 2016.
[12] H. S. Midtiby, S. K. Mathiassen, K. J. Andersson, and R. N. Jørgensen,
“Performance evaluation of a crop/weed discriminating microsprayer,”
Comput. Electron. Agric., vol. 77, no. 1, pp. 35–40, Jun. 2011.
[13] P. Lottes, M. Hoeferlin, S. Sander, M. M¨uter, and others, “An effective
classification system for separating sugar beets and weeds for precision
farming applications,” Proceedings of the, 2016.
[14] M. Weis and R. Gerhards, “Feature extraction for the identification
of weed species in digital images for the purpose of site-specific
weed control,” in Precision agriculture’07. Papers presented at the 6th
European Conference on Precision Agriculture, Skiathos, Greece, 3-6
June, 2007., 2007, pp. 537–544.
[15] H. T. Søgaard, “Weed classification by active shape models,” Biosystems
Eng., vol. 91, no. 3, pp. 271–281, 2005.
[16] M. Dyrmann, R. N. Jørgensen, and H. S. Midtiby, “RoboWeedSupport
- detection of weed locations in leaf occluded cereal crops using a
fully convolutional neural network,” in ECPA 2017 - 11th European
Conference on Precision Agriculture, 2017, submitted.
[17] Y. Zhang, E. S. Staab, D. C. Slaughter, D. K. Giles, and D. Downey,
“Automated weed control in organic row crops using hyperspectral
species identification and thermal micro-dosing,” Crop Prot., vol. 41,
pp. 96–105, Nov. 2012.
[18] P. Rydahl, “A danish decision support system for integrated management
of weeds,” Aspect.s of Applied Biology, vol. 72, 2004.
[19] L. N. Jørgensen, E. Noe, A. M. Langvad, J. E. Jensen, and others,
“Decision support systems: barriers and farmers’ need for support,”
EPPO, 2007.
[20] M. Sønderskov, P. Kudsk, S. K. Mathiassen, O. M. Bøjer, and P. Rydahl,
“Decision support system for optimized herbicide dose in spring barley,”
Weed Technol., vol. 28, no. 1, pp. 19–27, 2014.
[21] J. M. Montull, M. Soenderskov, P. Rydahl, O. M. Boejer, and
A. Taberner, “Four years validation of decision support optimising
herbicide dose in cereals under spanish conditions,” Crop Prot., vol. 64,
pp. 110–114, Oct. 2014.
[22] M. Sønderskov, R. Fritzsche, F. de Mol, B. Gerowitt, S. Goltermann,
R. Kierzek, R. Krawczyk, O. M. Bøjer, and P. Rydahl, “DSSHerbicide:
Weed control in winter wheat with a decision support system in three
south baltic regions – field experimental results,” Crop Prot., vol. 76,
pp. 15–23, Oct. 2015.
[23] T. Been, A. Berti, N. Evans, D. Gouache, V. Gutsche, J. E. Jensen,
J. Kapsa, N. Levay, N. Munier-Jolain, S. Nibouche, M. Raynal,
and P. Rydahl, “Review of new technologies critical to effective
implementation of decision support systems (DSSs) and farm
management systems (FMSs),” Aarhus University, Tech. Rep., 6 Mar.
2009.
[24] P. Rydahl, N.-P. Jensen, M. Dyrmann, P. H. Nielsen, and R. N.
Jørgensen, “RoboWeedSupport - presentation of a cloud based system
bridging the gap between in-field weed inspections and decision support
systems,” in ECPA 2017 - 11th European Conference on Precision
Agriculture, 2017, submitted.
[25] S. L. Madsen, M. S. Laursen, R. N. Poulsen, and R. N. Jørgensen,
“RoboWeedSupport - semi-automated UAS system for cost efficient
high resolution i sub millimeter scale acquisition of weed images,” in
ECPA 2017 - 11th European Conference on Precision Agriculture, 2017,
submitted.
[26] R. N. Jørgensen, M. S. Laursen, M. Dymann, and R. N. Poulsen,
“RoboWeedSupport - weed mapping with drones using a DJI phantom
4,” Denmark, 11 Oct. 2016.