Loading...
Please wait, while we are loading the content...
Similar Documents
An Unsupervised Segmentation with an Adaptive Number of Clusters Using Cloude-Pottier Decomposition, Wishart Test Statistic and MCCV A lgorithm for Fully Polarimetric Sar Data Analysis
| Content Provider | Semantic Scholar |
|---|---|
| Author | Cao, Fang Wu, Yirong Pottier, Eric |
| Copyright Year | 2007 |
| Abstract | In this paper, an unsupervised segmentation method is proposed for fully polarimetric SAR data. We introduce the MCCV algorithm to automatically estimate the optimal number of clusters from the data inner structure. The experimental results show that the number of classes is a crucial point for the unsupervised segmentation, which will strictly affect the segmentation performance. The MCCV algorithm represents to be a reliable method to estimate the appropriate number of clusters, and the proposed segmentation algorithm also provides better performance than the general Wishart H/α/A segmentation. |
| Starting Page | 48 |
| Ending Page | 48 |
| Page Count | 1 |
| File Format | PDF HTM / HTML |
| Volume Number | 644 |
| Alternate Webpage(s) | http://earth.esa.int/workshops/polinsar2007/presentations/261_cao.pdf |
| Alternate Webpage(s) | http://earth.esa.int/workshops/polinsar2007/papers/261_cao.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |