home    about    browse    search    latest    help 
Login | Create Account

Digital images for assessing soil cover of crop plants

Rasmussen, Jesper and Nørremark, Michael (2007) Digital images for assessing soil cover of crop plants. Speech at: 7th EWRS Workshop on Physical and Cultural Weed Control, Salem, Germany, 11-14 March 2007.

[img] PDF - English


The main drawback by using crop soil cover in weed harrowing research is that it is assessed by visual scores, which are biased and context dependent. This problem may be solved by using digital image analysis. In this paper a new image capture standard and digital image analysis procedure was used to illustrate three key issues in relation to weed harrowing; selectivity, resistance and recovery. All issues require reliable assessments of crop soil cover. Crop soil cover was deduced from assessments of leaf cover, which is defined as the proportion of pixels in digital images determined to be green.
Objective assessments of leaf cover and crop soil cover showed that the selectivity of weed harrowing in winter wheat and spring barley was unaffected by timing within a two weeks interval. Crop recovery, defined as the ability of the crop to recover from soil cover was determined in winter wheat and highly influenced by timing of weed harrowing. Increasing intensities of harrowing in growth stage (BBCH) 22 in winter wheat increased crop yields whereas crop yields declined by increasing intensities in growth stage 23 due to differences in the crop recovery capacity. Resistance defined as the capacity of the crop to resist soil cover was tested in barley, field pea and mixtures of barley and field pea. No differences were found between the crops. Future aims in mechanical weed control research are discussed in the context of the availability of unbiased crop soil cover data.

EPrint Type:Conference paper, poster, etc.
Type of presentation:Speech
Subjects: Crop husbandry > Weed management
Research affiliation: Denmark > KU - University of Copenhagen
Deposited By: Rasmussen, Associate professor Jesper
ID Code:10759
Deposited On:07 May 2007
Last Modified:12 Apr 2010 07:35
Document Language:English
Refereed:Not peer-reviewed

Repository Staff Only: item control page