The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLII-2/W16
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W16, 111–118, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W16-111-2019
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W16, 111–118, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W16-111-2019

  17 Sep 2019

17 Sep 2019

STATISTICAL ANALYSIS OF AIRBORNE IMAGERY COMBINED WITH GIS INFORMATION FOR TRAINING DATA GENERATION

G. Häufel1, D. Bulatov1, and P. Helmholz2 G. Häufel et al.
  • 1Fraunhofer IOSB, Gutleuthausstr. 1, 76275 Ettlingen, Germany
  • 2Spatial Sciences, Curtin University, GPO Box U1987, Perth, WA 6845, Australia

Keywords: training data generation, GIS data, Mean-Shift, K-Means, classification, land cover

Abstract. In recent years, the task of land cover classification from airborne image and elevation data advanced considerably due to enhanced applicability of CNNs (Convolutional Neural Networks). Nevertheless, CNNs require a huge amount of training data. Traditionally, few essential feature values, such as elevation or vegetation index, had been chosen to provide a coarse distinction of classes, but very often these values have to be adapted depending on the imagery. To improve this process, freely available GIS data are combined with spectral and spatial features (and their variations) following the K-Means and Mean-Shift algorithm. Based on cluster assignments to pixels, statistical analysis for extracting plausible values for distinguishing between land cover classes is applied. The resulting labeled databases are evaluated using ground truth data, and will form the basis for the training data required for CNNs.