| Creator |
|
|
|
| Language |
|
| Publisher |
|
|
|
| Date |
|
| Source Title |
|
| Vol |
|
| Publication Type |
|
| Access Rights |
|
| Related DOI |
|
|
|
|
|
| Related URI |
|
|
|
|
|
| Relation |
|
|
|
|
|
| Abstract |
We consider supervised image classification based on Markov Random Fields (MRFs), which are derived by the Jeffreys divergence between class-conditional probability densities. It is shown that the MRF... gives an extension of Switzer’s smoothing method for classification. Further, the exact error rate due to the MRF is obtained in a simple setup, and several properties of the classification are derived. Our procedure is applied to two multispectral images, and shows a good performance.show more
|
Export Mendeley
2004-17