Volume XLII-2/W6
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W6, 183-188, 2017
https://doi.org/10.5194/isprs-archives-XLII-2-W6-183-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W6, 183-188, 2017
https://doi.org/10.5194/isprs-archives-XLII-2-W6-183-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.

  23 Aug 2017

23 Aug 2017

DEVELOPMENT OF A ROBUST IMAGE MOSAICKING METHOD FOR SMALL UNMANNED AERIAL VEHICLE

J. Kim and T. Kim J. Kim and T. Kim
  • Dept. of Geoinformatic Engineering, Inha University, Incheon, Korea

Keywords: UAV, Image Mosaicking, Imaging Geometry Characteristics, Near-ortho mosaic

Abstract. In this paper, a tie-point based image mosaicking method considering imaging characteristics of small UAVs is proposed. Small UAVs can be characterized to have unstable flight trajectory and lower flight height. The proposed method considers the imaging characteristics in image transformation estimation and image blending process. For image transformation estimation, an optimal transformation model is variably applied by using tie-point area ratio. The optimal tie-point area ratio was about 0.3. Mosaicking error was largely decreased by using this tie-point area ratio. For image blending, a composite area minimization is introduced as a preceding step of image resampling. Composite areas of individual images were minimized by analyzing image overlaps between adjacent images. The proposed method was evaluated over flat area and urban area with highly overlapping multi-strip and inconsistently overlapping strip. Experiment results showed that the proposed method can reliably generate mosaics not only from UAV images acquired in good environment but also from extreme environment.