基于变化检测的滨海湿地图高效更新方法

Efficient method for updating coastal wetland map based on change detection technology

目前对变化剧烈的海岸带区域的遥感监测仍依赖于效率较低的人工提取,其结果现势性较差,且是对整幅图的重新提取,无法满足管理部门对海岸带监测的现势性要求。针对该问题,提出了基于遥感图像变化检测的滨海湿地专题图高效更新方法。首先采用差值-主成分分析方法获取监测时段内发生变化的滨海湿地区域;然后采用决策树分类方法对变化区域进行分类;最后利用变化区域的分类结果更新历史专题图,实现对已有专题图的更新。以更新辽宁省双台河口国家级自然保护区滨海湿地专题图为例,证实该文提出的方法高效、准确,且易于操作,具有在滨海湿地资源调查中推广的价值。

Nowadays, the method for detecting coastal wetlands by remote sensing still depends on manual extraction, and its long cycle makes the results not current; moreover, in most time, people must extract the whole map again, so the result can＇ t satisfy the needs for management. To solve this problem, this paper proposed an efficient method for updating coastal thematic map based on the change detection technology by using remote sensing images. First, the difference and principal component method is used to get the change area; second, the change areas are classified by using the decision tree method; third, the old thematic map is updated based on the classification result of the change area, and therefore the later thematic map is obtained. In this paper, the authors chose the Shuangtai River Mouth National Nature Reserve as the study area, and the research result indicates that the method proposed in this paper is efficient, accurate and characterized by a simple process, thus deserving extension in resources investigation of coastal wetlands.