Acquiring plant features with optical sensing devices in an organic strip-cropping system
Krus, Anne ; Apeldoorn, Dirk Van; Valero, Constantino ; Ramirez, Juan José - \ 2020
Agronomy 10 (2020)2. - ISSN 2073-4395
Cabbages - Lidar - Plant extraction - Point cloud - Weighted sum
The SUREVEG project focuses on improvement of biodiversity and soil fertility in organic agriculture through strip-cropping systems. To counter the additional workforce a robotic tool is proposed. Within the project, a modular proof of concept (POC) version will be produced that will combine detection technologies with actuation on a single-plant level in the form of a robotic arm. This article focuses on the detection of crop characteristics through point clouds obtained with two lidars. Segregation in soil and plants was successfully achieved without the use of additional data from other sensor types, by calculating weighted sums, resulting in a dynamically obtained threshold criterion. This method was able to extract the vegetation from the point cloud in strips with varying vegetation coverage and sizes. The resulting vegetation clouds were compared to drone imagery, to prove they perfectly matched all green areas in said image. By dividing the remaining clouds of overlapping plants by means of the nominal planting distance, the number of plants, their volumes, and thereby the expected yields per row could be determined.
Extrapolation of in situ data from 1-km squares to adjacent squares using remote sensed imagery and airborne lidar data for the assessment of habitat diversity and extent
Lang, Mait ; Vain, R. ; Bunce, R.G.H. ; Jongman, R.H.G. - \ 2015
Environmental Monitoring and Assessment 187 (2015). - ISSN 0167-6369 - 16 p.
Plant life forms - General habitat categories - Lidar - Landsat-7 Enhanced ThematicMapper Plus - Iterative self organising clustering - Maximumlikelihood
Habitat surveillance and subsequent monitoring
at a national level is usually carried out by recording
data from in situ sample sites located according to
predefined strata. This paper describes the application
of remote sensing to the extension of such field data
recorded in 1-km squares to adjacent squares, in order to
increase sample number without further field visits.
Habitats were mapped in eight central squares in northeast
Estonia in 2010 using a standardized recording
procedure. Around one of the squares, a special study
site was established which consisted of the central
square and eight surrounding squares. A Landsat-7
Enhanced Thematic Mapper Plus (ETM+) image was
used for correlation with in situ data. An airborne light
detection and ranging (lidar) vegetation height map was
also included in the classification. A series of tests were
carried out by including the lidar data and contrasting
analytical techniques, which are described in detail in
the paper. Training accuracy in the central square varied
from 75 to 100 %. In the extrapolation procedure to the
surrounding squares, accuracy varied from 53.1 to
63.1 %, which improved by 10 % with the inclusion
of lidar data. The reasons for this relatively low classification
accuracy were mainly inherent variability in the
spectral signatures of habitats but also differences between
the dates of imagery acquisition and field sampling.
Improvements could therefore be made by better
synchronization of the field survey and image acquisition
as well as by dividing general habitat categories
(GHCs) into units which are more likely to have similar
spectral signatures. However, the increase in the number
of sample kilometre squares compensates for the loss of
accuracy in the measurements of individual squares.
The methodology can be applied in other studies as
the procedures used are readily available.