高分遥感影像林地资源信息提取方法研究

Information Extraction Method of the Land Forest Resource through Remote Sensing Image of High Resolution

林地资源是森林资源的重要组成部分,在空气净化、自然防疫、制氧、气候调节等方面均具有重要作用。本文应用GF-1卫星数据,基于e Cognition软件平台,采用面向对象方法对廊坊市永清县2013年11月GF-1影像进行林地信息提取。通过多尺度分割试验确定最优尺度分割参数,结合影像地物纹理、光谱特征,采用模糊逻辑分类算法实现林地信息高效提取。

Forest land is an important part of the forest resources, which plays a key role in air purification, natural epidemic preven- tion, oxygen production, climate regulation and so on. Based on the eCognition software platform, the article explores GF-1 satellite data and uses the object-oriented method to conduct the information extraction of the forest land from the GF-1 image of Yongqing County, Langfang city in November, 2013. Through the multi-scale segmentation tests, the optimal scale parameter is determined. And the information extraction of the forest land can be achieved by the fuzzy logic classification algorithm and the image spatial texture and spectral characteristics.