Measurement of soil tillage using uav high-resolution 3d data

Carla Rebelo, João Nascimento

Research output: Contribution to journalArticlepeer-review


Remote sensing methodologies could contribute to a more sustainable agriculture, such as monitoring soil preparation for cultivation, which should be done properly, according to the topographic characteristics and the crop’s nature. The objectives of this work are to (1) demonstrate the potential of unmanned aerial vehicle (UAV) technology in the acquisition of 3D data before and after soil tillage, for the quantification of mobilised soil volume; (2) propose a methodology that enables the co-registration of multi-temporal DTMs that were obtained from UAV surveys; and (3) show the relevance of quality control and positional accuracy assessment in processing and results. An unchanged-area-matching method based on multiple linear regression analysis was implemented to reduce the deviation between the Digital Terrain Models (DTMs) to calculate a more reliable mobilised soil volume. The production of DTMs followed the usual photogrammetric-based Structure from Motion (SfM) workflow; the extraction of fill and cut areas was made through raster spatial modelling and statistical tools to support the analysis. Results highlight that the quality of the differential DTM should be ensured for a reliable estimation of areas and mobilised soil volume. This study is a contribution to the use of multi-temporal DTMs produced from different UAV surveys. Furthermore, it demonstrates the potential of UAV data in the understanding of soil variability within precision agriculture.
Original languageEnglish
Article number4336
Pages (from-to)1-28
Number of pages28
JournalRemote Sensing
Issue number21
Publication statusPublished - 1 Nov 2021


  • Co-registration
  • Differential DTM
  • Digital terrain model
  • Positional accuracy
  • Precision agriculture
  • SfM
  • Soil tillage
  • Volume changes


Dive into the research topics of 'Measurement of soil tillage using uav high-resolution 3d data'. Together they form a unique fingerprint.

Cite this