利用边缘检测对高分辨率影像中建筑物的提取研究

Research on Building Extraction from High Resolution Remote Sensing Images with Edge Detection

遥感作为一种探测技术,越来越广泛地运用到城市建设等应用研究领域中。近年来,高分辨率遥感影像的应用推动了城市遥感进一步向定量化方向发展,同时,也大大增加了数据处理的工作量。如何从海量的遥感数据中筛选出有价值的信息一直是遥感影像处理工作最需要解决的难题。将尝试使用Sobel算子与数学形态学相结合的方法,利用2001年阜新市区的QuickBrid多光谱影像,探索一种高分辨率多光谱影像建筑物的提取方法。

As the detection technology,remote sensing technology has been increasingly used in the field of urban construction and other related research fields. In recent years, the development of high resolution remote sensing promotes urban remote sensing to have a further quantitative development and greatly increases data-processing workload;how to select the valuable information from the mass of remote sensing data has been a resolved problem. A method was put forward based on Sobel operator and mathematical morphology. Adopting a QuickBird multi-spectral image of Fuxin District in 2001, the method of building extraction from high resolution and multi-spectral images was proposed.