Texture analysis is a fundamental aspaect of discriminating distinct regins in remotely sensed imagery.Features derived from cooccurrence matries are a popular method for identifying ice and open water classes in SAR sea ice imagery.Gabor filters offer another approach for such interpretation.These filters are based on characteristics of the human visual system and are able to generate a multi-scale decomposition of a signal.here,the images are segmented in an unsupervised fashion using only texturally based features (ie.tone is excluded during the feature extraction process).
展开▼