we will use the split and merge techniques based on high
level features such as context and semantic features.
References
Akçay, H.G., and S. Aksoy, 2008. Automatic detection of geospatial
objects using multiple hierarchical segmentations,
IEEE Trans-
actions on Geoscience and Remote Sensing
, 46(7):2097–2111.
Baatz, M., and A. Schape, 2000. Multi-resolution segmentation:
an optimization approach for high quality Multiscale
image segmentation,
Angewandte Geographische
Informationsverarbeitung XII
(J. Strobl, editor), Wichmann,
Heidelberg, pp. 12–23.
Battiti, R., 1994. Using mutual information for selecting features in
supervised neural net learning,
IEEE Transactions on Neural
Networks
, 5(4):537–550.
(a)
(b)
(c)
Figure 10. Results of K-means clustering algorithm on DS7 with mask 3: (a) panchromatic image, (b) clustering with the proposed FS
method, and (c) clustering with Relief method.
(a)
(b)
(c)
(d)
Figure 11. Results of K-means clustering algorithm on DS4 with mask 3: (a) panchromatic image, (b) clustering with the proposed FS
method, (c) clustering with MRMR method, and (d) clustering with CMIM method.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
March 2016
221
