论文部分内容阅读
城市变化监控能够有效帮助政府机关和城区规划部门更新地图并制定长远的发展计划。论文提出了一个基于Log-Gabor滤波器快速提取遥感图像中居民区域的新方法。方法分为三步:首先使用Log-Gabor滤波器提取遥感图像中具有边缘方向性的城市居民区域特征;然后在此基础上“耦合”相同居民区域不同方向滤波器提取出来的纹理特征,进而获得完整的居民区域;最后利用Otsu方法在结果图像中对居民区域进行标记和分割。实验将本论文方法同传统纹理提取手段共生矩阵分析方法进行比较证明了本文方法的优越性。
City change monitoring can effectively help government agencies and urban planning departments update maps and develop long-term development plans. This paper presents a new method based on Log-Gabor filter to quickly extract the residential area in remote sensing images. The method is divided into three steps: firstly, the Log-Gabor filter is used to extract the regional characteristics of urban residents with edge directionality in the remote sensing images; secondly, the texture features extracted from different direction filters in the same residential area are “coupled” And then get a complete residential area; Finally, using Otsu method in the results of the image of the residential area marking and segmentation. The experiment compares the method of this thesis with that of the traditional texture extraction method, and shows the superiority of this method.