The classification process was divided into the following steps: 1) input images, 2) multi-resolution segmentation, 3) image object hierarchy, 4) creation of class hierarchy, 5) classification using training samples and standard nearest neighbor, 6) classification base on segmentation, 7) repetition of steps for best result, and 8) final merge classification (Laliberte et al. 2004).
Because the signatures of permanent and shifting cultivation are similar, except at the start of the year when vegetation is sometimes more dense for shifting cultivation (i.e., the vegetation has not been cleared yet), the two land cover classes were merged post-classification under "agricultural land" and reported as such.
The polygon-based format helps to edit or merge terrain classification data easily.
The use of synonym merging improved classification accuracy by approximately 3 percentage points compared to use CHI only.
Then, 8 groups of TUs were formed by merging both classification systems.
For reads containing more than one EGT and therefore more than one classification, the taxonomic classifications were merged into a single one that was determined as the lowest common ancestor of all classifications.
The frequent presence of clouds and Landsat7 scan line gaps forced us to conduct of series of separate classifications in each year, which were then merged by choosing whichever classification used the highest number of cloud- and gap-free bands at any given pixel.
To get a more general view of the categories and the classification of the terms, all previously mentioned classifications are merged into a comprehensive multidimensional view (see Figure 6).
To estimate the appropriate number of clusters, the similar clusters should be merged to given the classification map with different number of clusters.
If the synonym are merged, not only the classification accuracy will be improved, but also the dimension of the feature space can be reduced so that the efficiency of the algorithm can be improved.
