Tidal Channel Information Extraction And Research Of Jiuduansha Based On Multi-source Remote Sensing Data

Tidal channel is one of the major landforms of tidal flat. And it is also animportant part of tidal flat. It is located in the active zone of intersection of sea andland. Tidal channel is the main road for material and energy exchange between tidalflat and open coastal water body. And it is also the place for ships sailing andanchorage. By the study of the tidal channel, scientific foundations can be providedfor the rational development and utilization of the tidal flat resources.Remote sensing technology has the advantages of macro, dynamic andsynchronous monitoring. By the research of remote sensing of tidal channel, scientificfoundations for the rational development and utilization of tidal flat resources couldbe provided and it also could make up for the deficiency of the common observationmethods. High resolution satellite-borne synthetic aperture radar technology, as anefficient earth observation method, has the ability to obtain ground information in allweather conditions and all time. The German X-Band TerraSAR-X satellite waslaunched on June15,2007. It is satellite-borne synthetic aperture radar with very highspatial resolution (up to one meter). In this paper, TerraSAR-X data was used toextract the tidal channel information in the Jiuduansha of Yangtze estuary. Thisresearch tried to find an efficient method to extract and quantify the information oftidal channel based on the microwave remote sensing characteristics of tidal channel.By using the method of two peaks threshold and mathematical morphology, wesucceed in extracting the tidal channel of Jiuduansha based on SPOT and TerraSAR-Ximage. In the course of the study, we got the following results:(1) The panchromatic band remote sensing data received by SPOT-5on March16,2010was used. Firstly, we analyze the histogram of Jiuduansha original image. Secondly, we find that two peaks threshold is appropriate for extracting the tidalchannel. Then, we find out applicable point for threshold. As results shown, thismethod has preferable applicability on the extraction of tidal channel information.(2) By using the common filter algorithm to deal with the noise of TerraSAR-Ximage. We compared mean value、deviation、ENL and ESI with pixels of before andafter filtering, and find out enhanced Lee filter is appropriate for filtering theTerraSAR-X image.(3)In this paper, the TerraSAR-X high resolution satellite-borne synthetic apertureradar data received on December25,2009was obtained, and the enhanced Lee filterwas used to denoise the TerraSAR-X image. Then the tidal channel of Jiuduansha wasextracted by mathematical morphology. As results shown, this method has preferableapplicability on the extraction of tidal channel information in high resolution radarimages.