Pre-harvest weed mapping of Cirsium arvense in wheat and barley with off-the-shelf UAVs

Rasmussen, Jesper; Nielsen, Jon; Streibig, J. C.; Jensen, Jens Erik; Pedersen, Ken Steen; Olsen, Søren I.

doi:10.1007/s11119-018-09625-7

Cited by 54 publications

(60 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The chosen software, however, did not affect the results. For larger areas, stitching of images to create ortho‐images would be preferable, but there is no reason to believe, that ortho‐images would give different competition parameters (Rasmussen et al ., ).…”

Section: Discussionmentioning

confidence: 97%

“…The automated image analysis used in this study is not biased, but it is not generally applicable because it detects all green vegetation (Rasmussen et al ., ). Algorithms have recently been developed to detect C. arvense in UAV imagery (Mink et al ., ; Rasmussen et al ., ), but these algorithms are not able to separate C. arvense from other green vegetation (Rasmussen et al ., ) or from other green vegetation at the same height as C. arvense (Mink et al ., ).…”

Section: Discussionmentioning

confidence: 99%

“…Due to the low plot number in each field, it was decided not to produce ortho‐images. A previous study shows that it is of minor importance to image analysis whether ortho‐images or individual nadir images are used (Rasmussen et al ., ).…”

Section: Methodsmentioning

confidence: 97%

“…As discussed in Rasmussen et al . (), it is far from trivial to detect C. arvense automatically in cereals based on image analysis. In consequence, this study applies visual scoring methods of C. arvense coverage based on UAV images.…”

Section: Introductionmentioning

confidence: 99%

“…All previous competition studies used traditional weed assessment methods based on sampling techniques, where C. arvense shoots were counted or biomass was harvested in small plots (1 m 2 ). Much work on the detection of C. arvense based on images from unmanned aerial vehicles (UAV) has been carried out (St ahl et al, 2013;Garcia-Ruiz et al, 2015;Olsen et al, 2017;Sørensen et al, 2017;Mink et al, 2018;Rasmussen et al, 2019;Forero et al, 2019) but methods to count overlapping C. arvense shoots are still lacking. As discussed in Rasmussen et al (2019), it is far from trivial to detect C. arvense automatically in cereals based on image analysis.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A novel approach to estimating the competitive ability of Cirsium arvense in cereals using unmanned aerial vehicle imagery

Rasmussen

Nielsen

2020

Weed Research

View full text Add to dashboard Cite

Summary Eight experiments were carried out in Denmark to determine the yield loss of spring barley due to Cirsium arvense in farmers' fields and to suggest and evaluate a novel approach for quantifying C. arvense infestation in large plots. Literature about the competitive ability of C. arvense is old, scattered and inconclusive, and existing models for estimating crop yield loss are based on data from North America. This study showed that C. arvense coverage could be quantified from unmanned aerial vehicle imagery using a manual image analysis procedure. This gave similar results as scoring the coverage. Yield loss of spring barley due to C. arvense infestation assessed at harvest was given by Y = 100·(1−exp(−0.00170·X)) where Y is the percentage of crop yield loss and X is the percentage of C. arvense coverage. The yield loss was much lower than estimates from models that have been developed in North America. It is speculated that the main reason for this is the later emergence of C. arvense than the crop due to lower soil temperatures in spring. Grain moisture increased linearly with C. arvense coverage: M = 0.0310·X where M is the proportional (%) increase in grain moisture and X is the proportion (%) of C. arvense coverage. Automated image analysis procedures are needed to estimate C. arvense coverage on field scales, and further experiments are needed to reveal whether the low competitive ability of C. arvense found in this study is representative for Northern Europe.

show abstract

Section: Discussionmentioning

confidence: 97%