|
|
|
Classification of Remotely Sensed Data using Gravitational Symbolic Clustering
2 Divisive Method
- Reduce the data by choosing an appropriate bin size and threshold. Let N be the initial number of samples in a sample set S. And let the number of clusters be equal to one, with N number of samples.
- Set a threshold value TCCSm.
- Using the similarity measure, find all the mutual pairs present in the symbolic data set.
- For all the Mutual Pairs present find the CCS 'CCSm between samples. Check if
CCSm>TCCSm, if so then split the Mutual pairs into two separate clusters. Increment the
number of clusters by 2.
- Step 3 and 4 is repeated for all the mutual pairs present at that stage.
- Determine the Cluster Indicator value for each Pth merging as in equation 6.:
- Decrement TCCSm in steps.Iterate steps 3 to 6 until a stage is reached when no replacement
of mutual pair occurs.
Results
Image of Moon taken by Galileo space craft on December 9 1990, with catalog number PIA00113 is taken for testing. Size of image is 535 * 535 with 3 features. Both Agglomerative and Divisive Gravitational Symbolic methods are considered, results are as shown in figure 1 and 2 respectively .
Plot of Cluster Index Vs Number of Clusters is also shown in figure 3.
Fig 1. Classification Map and details of Classification Cover of Moon (Agglomerative method)
Fig 2. Classification Map and details of Classification Cover of Moon (Divisive method)
Fig 3. Variation for CI for Moon Image ( Series 1 Agglomerative Method Series 2 Divisive metod.)
|
|
|