Abstract:

A texture features extraction method based on Worldview-II multi-spectral remote sensing data is proposed. Aiming at the eight bands of Worldview-II, the multispectral remote sensing data are compressed by principle component analysis(PCA) conversion. The principle component gray level difference frequency space (GLDFS) is established according to the compressed data. And then the texture features are extracted using GLDFS. Experimental results show that the proposed method can retain the coordination relationship among multi-spectral remote sensing data. Compared with the traditional single band texture analysis method based on gray level co-occurrence matrix (GLCM), the proposed method has higher classification precision and efficiency