遥感图像纹理信息提取方法综述

THE STUDY PROGRESS OF THE TEXTURE FEATURE EXTRATION FOR REMOTE SENSING IMAGE

纹理是遥感图像的重要特征,它提示了图像中辐射亮度值空间变化的重要信息。要利用图像空间信息提高分类精度,合理而有效地度量纹理至关重要。目前遥感图像纹理信息提取方法主要有:统计描述法、小波变换法、分维分形法和地统计学4类。分别就各种方法的优缺点、适用领域和应用情况进行了阐述,最后展望了遥感图像纹理信息提取方法的发展方向和研究热点。

The texture is one of the important features of remote sensing images, which is related to the spatial distribution of the intensity value in the image and as such contains information regarding contrast, coarseness, directionality, ect. A considerable number of quantitative texture features can be extracted from images using different methodologies in order to characterize these properties, and then can be used to classify pixels following analogous processes as with spectral classifications. The extraction of texture features from high resolution remote sensing imagery provides a complementary source of data for those applications in which the spectral information is not sufficient for identification or classification of spectrally heterogeneous landscape units. In order to increase the accuracy of classification, it is very vital to describe image texture reasonably and effectively and to choose the suitable texture modeling. At present, there is a wide range of texture analysis techniques that are used for feature extraction： Statistical methods （grey level occurrence matrix, semivariogram analysis） ; structure methods, modeling methods and filter techniques （energy filters, Gabor filters） . The combination of parameters that optimize a method for a specific application should be decided when these techniques are used. These parameters include the neighbourhood size, the distance between pixels, the type of filter or mother wavelet used, the frequency or the standard deviation used to create the Gabor filters, etc. The combination of parameters and the texture method used is expected to be key in the success and efficiency of these techniques for a particular application. Recently, Geo-statistic methods and Wavelet decomposition methods are both applied widely and have a great effect on accurate rate of remote sensing image classification.