摘要
随着城市化进程的加快,如何准确、高速、有效的从测绘数据中提取建筑物信息变得尤为重要。目前建筑物分类提取后,建筑物边缘不清晰,局部破碎,地物类别错分形成小斑点等一系列问题仍然存在。通过数学形态学和计算机数字图像处理的一些算法,利用MATLAB编程,将小斑点去除,修复建筑物内部破碎区域,清晰建筑物边缘,最后通过canny算法对建筑物边缘进行提取,结果表明本文对建筑物分类提取后细化处理具有较好的可行性与有效性。
With the development of urbanization,it becomes much more important to extract the buildings＇ features from data exactly,quickly and efficiently.After extracting,the edges of the buildings are not quite clear,some fractions are partly left,and many small spots caused by the misclassification of features still remain.Here,by using the algorithms of mathematical morphology and digital image processing,and programming with MATLAB,the small spots would be removed,the fractured parts of inner buildings would be restored,and the edges of buildings would be made clear.Finally,by using the algorithm of canny,the edges of buildings would be extracted,which could prove the feasibility and validity of image processing after building extraction.
出处
《水利与建筑工程学报》
2012年第1期105-108,共4页
Journal of Water Resources and Architectural Engineering
基金
国家自然科学基金资助项目(51079053)
关键词
建筑物提取
数学形态学
腐蚀
膨胀
小面积去除
building extraction
mathematical morphology
erosion
dilation
small-sized removal
