基于形态学空间特征的高光谱遥感图像分类方法

Classification of hyperspectral remotely sensed images based on the extraction of spatial feature

传统的高光谱图像分类主要是基于像素的光谱特征,在一定程度上忽略了高光谱遥感图像中像素之间的空间相关性。为了充分利用高光谱图像中的空间信息,提出了一种基于加权多结构元素无偏差形态学的空间特征提取方法,并基于形态学的多尺度特征和结构保持性提出了基于邻域的多尺度空间特征提取方法,得到了高光谱遥感图像的空间特征。对k-NN分类算法进行改进,提出了基于变精度粗糙集和重构误差的k-NN分类算法,实现了基于空间特征的高光谱遥感图像分类。在两个不同的高光谱遥感图像的实验验证了基于空间特征和改进k-NN分类算法的性能。

Traditional classification methods for hyperspectral remotely sensed images are mostly based on the charac- ter of spectral feature. The spatial features of hyperspectral images are somehow neglected. To make a better use of the spatial information of hyperspectral images, a spatial feature extraction method is proposed based on weighted multi- structure elements mathematical morphological processing without bias. Another spatial feature extraction method based on neighborhood based on the multiscale character of morphological structure is proposed. Then the spatial feature of hy- perspectral is formed. An improved k-NN classification method is proposed based on variable precision rough sets and re- construction error and the method is used to classify hyperspectral image. Experiment results show that the proposed method can classify hyperspectral images successfully.