In this paper we applied the fuzzy ARTMAP algorithm for combining multispectral aerial
imagery and lidar data so that the individual strengths of each data source can
compensate for the weakness of the other. Test data from four different study areas with
different characteristics have been used. First, we filtered the lidar point clouds to
generate a Digital Terrain Model (DTM), and then the Digital Surface Model (DSM) and
the Normalized Digital Surface Model (nDSM) were generated. After that, we have
derived 22 attributes from both aerial image and lidar data by a number of algorithms.
The attributes include those derived from the Grey Level Co-occurrence Matrix (GLCM),
Normalized Difference Vegetation Indices (NDVI) and slope. Finally, a Fuzzy ARTMAP
was used to detect buildings, trees, roads and grass from the aerial image, lidar data and
the generated attributes.
The rules for tuning the parameters of the Fuzzy ARTMAP and its relation to the
produced classification accuracy have been studied. The ability of the Fuzzy ARTMAP
to detect features has been evaluated and compared against other classifiers. Also, the
transferability of information from one data set to another has been tested. Finally, the
contributions of the individual attributes to the quality of the classification results were
assessed. |
