综合Span图和纹理特征的高分三号影像建筑区提取

建筑区的识别和提取是城市环境规划与研究至关重要的工作。本文采用高分三号全极化SAR影像,提出了一种综合Span图和纹理特征的建筑区提取方法。首先基于Span图利用灰度共生矩阵算法提取图像的7种原始纹理特征,通过目视解译选择出4种纹理效果较好的统计量,然后利用主成分分析法去除他们之间的相关性,筛选出2个最佳纹理特征与Span图结合,最后对组合影像进行分类提取。本文将提取结果与综合灰度和纹理特征建筑区提取、无纹理特征提取方法结果进行对比,实验结果表明:本文方法提取建筑区边界轮廓更加清晰,精度可达92%,提取效果明显得到了优化。

Building Areas Extraction in GF-3 Images Based on the Integration of Span Image and Texture Features

The identification and extraction of building areas is a vital task in urban environmental planning and research. A novel method for building areas extraction using both Span map and texture features that based on the full polarization SAR image of GF-3 is proposed in this paper. Firstly, seven original texture features based on total power Span map are extracted by using GLCM. Secondly, four kinds of good texture statistics are selected by visual interpretation. Then, two best texture features are selected as the best texture statistics based on the principal component analysis method and combined with the Span image. Finally, the new image is classified and the building areas are extracted. In this paper, the extraction results are compared with the integrated grayscale and texture feature building areas extraction method and the non-texture feature extraction method. The experimental results show that the proposed method extracts the boundary contour of the building area more clearly, the accuracy of this paper is as high as 92%, and the extraction effect is obviously optimized.