Texture Feature Extraction Based on Millimeter-Wave SAR Sea Ice Data

Millimeter-wave synthetic aperture radar (SAR) technology has important applications in sea ice monitoring and observation. In order to better utilize the texture information in the millimeter-wave SAR sea ice data, the gray-level covariance matrix (GLCM) method is used to extract texture features. First, the millimeter-wave SAR sea ice data are preprocessed, including denoising and radiometric correction. Second, the GLCM method is used to calculate the gray-scale relationship statistics between pixel points in the sea ice image. By defining specific distances and angles, a GLCM matrix is obtained, which contains the statistics of different texture features in the sea ice image. Based on the GLCM matrix, several texture feature parameters such as energy, contrast, correlation and entropy are extracted, which describe the texture structure and the degree of change of the sea ice image, and detailed information about the texture features of sea ice can be obtained by analyzing the values and spatial distribution of these feature parameters. The experimental results show that the GLCM method can more accurately capture the texture details and changes in sea ice images.