基于GF-1多时相波段组合的武汉市土地利用变化检测

Change Detection of Land Use in Wuhan City Based on Multi Temporal Band Combination of GF-1

GF-1号卫星是中国自主研发的第一颗高分辨率遥感卫星。为分析该卫星影像数据在土地利用变化检测中的应用,本文以武汉主城区为研究对象,选取该地区2014年5月、2015年4月GF-1号卫星遥感影像,采用分类后比较与多时相波段组合直接分类两种方法分别进行试验,将二者实验结果比较分析。结果表明,多时相波段组合直接分类是一种更为有效的变化检测方法。相比分类后比较该方法不仅检测精度提高了将近17%,土地利用变化检测效果也明显改善。另外,对武汉主城区主要土地利用变化类型进行面积统计分析,对武汉主城区建设具有一定的指导意义。

GF-1 is the first high-resolution remote sensing satellite developed by China.Applying 2014 and 2015 GF-1 images as data sources of Wuhan city,the change detection of land use is analyzed.By using the post-classification and multi temporal band combination direct classification method,this paper compares the results of classification respectively.The results show that multi temporal band combination direct classification is a more effective method to detect the change of land use.Compared to the post classification,the detection accuracy increased nearly 17% and the detection effect of the change of land use is improved obviously via the method of multi temporal band combination direct classification.In addition,this paper statistics the areas of the main types of land use change and analyze the reasons finally,which has a certain guiding significance on the construction of Wuhan urban area.