The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLII-2/W18
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W18, 159–165, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W18-159-2019
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W18, 159–165, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W18-159-2019

  29 Nov 2019

29 Nov 2019

INVESTIGATING THE USABILITY OF UAV OBTAINED MULTISPECTRAL IMAGERY IN TREE SPECIES SEGMENTATION

M. Pap1, S. Kiraly2, and S. Moljak3 M. Pap et al.
  • 1IoT Research Center, Eszterházy Károly University, Eger, Hungary
  • 2Information Technology Department, Eszterházy Károly University, Eger, Hungary
  • 3Innoregio Knowledge Center, Eszterházy Károly University, Eger, Hungary

Keywords: UAV, image segmentation, image processing, multispectral imagery, energy plants

Abstract. A widely used form of renewable energy are bioenergy crops. One form of it is the energy forestry that includes short rotation coppice plantations in which fast growing species of tree or woody shrub are grown (e.g. poplar, willow). The accurate prediction of forest biomass and volume can be used for the evaluation of plant breeding efficiency as well. The automatic tracking of plant development by traditional methods is quite difficult and labor intensive. Since energy forestries often contain different trees for estimating their volume it is essential to find segments containing the same tree species in the image.

We investigated the applicability of a low cost UAV and an intermediate cost UAV in the field of agricultural image segmentation that is the first stage of biomass estimation (Gatziolis et al., 2015, Gaulton et al., 2015).

This paper is a case study that shows the results of several segmentation algorithms applied on imagery obtained by a low cost UAV with low-cost camera, and imagery gathered by a UAV and camera set that are of higher quality and price. In the case study, we have observed two small forestry areas that contained six different tree species and their hybrids. Our results show that more expensive, better-equipped drone shots do not necessary provide significantly better segmentation.