Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3, 205-210, 2014
https://doi.org/10.5194/isprsarchives-XL-3-205-2014
© Author(s) 2014. This work is distributed under
the Creative Commons Attribution 3.0 License.
 
11 Aug 2014
Geometric processing workflow for vertical and oblique hyperspectral frame images collected using UAV
L. Markelin, E. Honkavaara, R. Näsi, K. Nurminen, and T. Hakala Finnish Geodetic Institute, Remote Sensing and Photogrammetry, Masala, Finland
Keywords: Structure-from-motion, Oblique, Hyperspectral, UAV, 3D point cloud Abstract. Remote sensing based on unmanned airborne vehicles (UAVs) is a rapidly developing field of technology. UAVs enable accurate, flexible, low-cost and multiangular measurements of 3D geometric, radiometric, and temporal properties of land and vegetation using various sensors. In this paper we present a geometric processing chain for multiangular measurement system that is designed for measuring object directional reflectance characteristics in a wavelength range of 400–900 nm. The technique is based on a novel, lightweight spectral camera designed for UAV use. The multiangular measurement is conducted by collecting vertical and oblique area-format spectral images. End products of the geometric processing are image exterior orientations, 3D point clouds and digital surface models (DSM). This data is needed for the radiometric processing chain that produces reflectance image mosaics and multiangular bidirectional reflectance factor (BRF) observations. The geometric processing workflow consists of the following three steps: (1) determining approximate image orientations using Visual Structure from Motion (VisualSFM) software, (2) calculating improved orientations and sensor calibration using a method based on self-calibrating bundle block adjustment (standard photogrammetric software) (this step is optional), and finally (3) creating dense 3D point clouds and DSMs using Photogrammetric Surface Reconstruction from Imagery (SURE) software that is based on semi-global-matching algorithm and it is capable of providing a point density corresponding to the pixel size of the image. We have tested the geometric processing workflow over various targets, including test fields, agricultural fields, lakes and complex 3D structures like forests.
Conference paper (PDF, 860 KB)


Citation: Markelin, L., Honkavaara, E., Näsi, R., Nurminen, K., and Hakala, T.: Geometric processing workflow for vertical and oblique hyperspectral frame images collected using UAV, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3, 205-210, https://doi.org/10.5194/isprsarchives-XL-3-205-2014, 2014.

BibTeX EndNote Reference Manager XML