Recognition of lithological units in airborne SAR images using new texture features

Abstract

Synthetic Aperture Radar (SAR) images are used increasingly in applications of remotely-sensed data for land-use investigation and in the recognition of geological features. This is because SAR images are rich in textural information and textural characteristics of an image are found to be useful in identification of lithological units. For this reason, the automatic processing and interpretation of SAR images are often focused on texture analysis, including the extraction of textural information, textural filtering (enhancement of textural perception and texture classification. This Letter evaluates a set of new texture features which are extracted from the texture spectrum. Some of the traditional texture features, extracted from the co-occurrence matrix, have been used for comparison with the new measures in the discrimination of different lithological units of an airborne SAR image. The results show the promising discriminating performance of the proposed features.