基于Pleiades-1卫星数据薇甘菊信息提取

Information extraction of Mikania micrantha based on Pleiades-1 data

薇甘菊Mikania micrantha是一种极具危险性的外来物种,对侵入地生态系统的完整性构成了极大威胁,及时、准确掌握薇甘菊分布信息成为防控其入侵的关键技术。目前,薇甘菊的监测主要采用人工方法,耗时费力。随着遥感技术的发展,采用高分辨率遥感数据快速准确提取薇甘菊分布信息成为可能。以深圳市Pleiades-1影像为主要信息源,开展薇甘菊信息的快速提取研究。结果表明:(1)Pleiades-1影像分辨率较高,在开展薇甘菊识别时的最佳波段组合是R(B3)、G(B2+B4)/2、B(B1);(2)通过Brovey变换、Gram-Schmidt变换、Pansharpen变换的对比分析,Gram-Schmidt变换在信息量、最佳波段、清晰度等方面,均优于Brovey变换和Pansharpen变换,既保留了大量的原有信息,又增强目视解译效果,是影像融合时的最优变换;(3)基于光谱与纹理信息分析,提取薇甘菊信息范围更好;(4)薇甘菊信息提取的最佳分割尺度是30,综合识别精度达95.3%,说明采用高分辨率遥感影像能实现薇甘菊信息的精细识别。

Mikania micrantha H.B.K. is a highly dangerous invasion plant, has seriously threaten the integrity of ecosystem of invaded area. It is required that timely and accuratly understand distribution ofM. micrantha information in China, and it is also very important for us to master and control M. micrantha intrusion. At present, the monitoring of M. micrantha adopt mainly artificial methods, take time and effort. With the development of remote sensing technology, using high resolution remote sensing data and quickly and accurately extracting M. micrantha distribution information become possible. By taking Shenzhen city＇s Pleiades-1 satellite images as the main source of information, the author of this study conducted the study on the fast extraction of M micrantha information. The results show that （1） the images had higher resolution, and the optimal band combinations that re-cognized ~/L micrantha are R（B3）,G（B2＋B4）/2 andB（B1）; （2） through comparing and analyzing Brovey transform, Gram-Schmidt transform and Pansharpen transform, we can find that Gram-Schmidt transformation in aspects such as information amount, best band and clarity etc. , are superior to Brovey transform and Pansharpen transform, it retains much of the original information, and enhance the effect of visual interpretation, is the optimal transformation for image fusion; （3） with the analysis methods of spectrum and texture information, we can get more wider extracting range for M. micrantha information; （4） the best segmentation scale for M. micrantha information extraction is 30, the integrated identification precision reached 95.3%. It is found that we can realize the precise identification of M. micrantha information by using high-resolution remote sensing images.