Use of texture features in discrimination of sea ice types in SAR images
Ice movement algorithms usually require a suitable number of high-contrast features to be present in two consecutive images. These features should be imaged in a consistent way. A set of texture features was tested to increase the contrast or separability between flat ice or ridged ice and open wate...
| Published in: | 10th Annual International Symposium on Geoscience and Remote Sensing |
|---|---|
| Main Author: | |
| Other Authors: | |
| Format: | Other Non-Article Part of Journal/Newspaper |
| Language: | English |
| Published: |
IEEE Institute of Electrical and Electronic Engineers
1990
|
| Subjects: | |
| Online Access: | https://cris.vtt.fi/en/publications/f43cb2b2-8a8b-4616-bacf-d53fd1503ecc https://doi.org/10.1109/IGARSS.1990.688978 |
| Summary: | Ice movement algorithms usually require a suitable number of high-contrast features to be present in two consecutive images. These features should be imaged in a consistent way. A set of texture features was tested to increase the contrast or separability between flat ice or ridged ice and open water or new thin ice (nilas). To quantify the improvement in the separability of the selected ice types, a classification of the SAR images was carried out separately for the original data and the texture image. |
|---|