一种融合数学形态学运算的多尺度建筑物分割算法

多尺度分割算法是面向对象的图像分析方法中的一种较为成熟、稳定的分割算法,但存在部分分割不准确和分割效率低等问题.为此,提出一种融合数学形态学运算的多尺度建筑物分割算法:首先利用高斯滤波器对遥感图像进行滤波处理,然后通过数学形态学运算对图像进行腐蚀处理,最后采用多尺度分割算法得到建筑物影像.对昆明市局部区域的QuickBird图像进行多尺度建筑物分割实验的结果表明,该算法具有分割精度高、效率快等优点,对类似工作有一定的借鉴意义.

A Building Segmentation Algorithm Based on Mathematical Morphology and Multi- scale Segmentation

Multi-scale segmentation approach is a relatively mature and stable algorithm in object-oriented image analysis algorithms;nevertheless,the problems of partial segmentation inaccuracy and low segmentation efficiency remain existent in this means.In view of this situation,the authors propose an algorithm combining the mathematical morphological operation with multi-scale building segmentation in this paper.First,the Gaussian filter is used to filter the image.And then,the mathematical morphological operation is used to erode the image.Finally,the multi-scale segmentation algorithm is applied to the image to extract the buildings.Experiments were conducted on the QuickBird image of a local area in Kunming,and the results show that the algorithm presented in this paper is more accurate and efficient than the traditional algorithms,and is thus of reference significance for similar segmentation work.