Font Size: a A A

Research On The Extraction Of Flow Pattern Using Remote Sensing Method In Inland Lake

Posted on:2019-04-16Degree:MasterType:Thesis
Country:ChinaCandidate:Y F ZhaoFull Text:PDF
GTID:2310330548457977Subject:Surveying and mapping engineering
Abstract/Summary:PDF Full Text Request
Lake flow is an important lake movement and one of the most important hydrological phenomena in the lake.It not only directly affects the transport and conversion of matter and energy in the lake,but also is related to the exchange and transfer of material and energy between the lake and the outside world.Therefore,mastering the fluid state information of lake is conducive to providing services such as forecasting,treatment and control for production,transportation,fishery,and port construction.At present,there are some commonly used methods,such as field observation,numerical simulation,multi-phase tracer detection,etc.,which generally have the disadvantages of high complexity,time-consuming operation,and inability to respond to changes in lake flow in time.In view of the above problems,this paper based on the existing research from the perspective of optical remote sensing image analysis carried out some research.In this paper,the inland lake-Poyang Lake as the research object,the characteristics of the lake in Poyang Lake were studied in depth.First of all,according to the existing research data of the lake and the characteristics of the Poyang Lake,the lake of Poyang Lake is classified,including the flow of invaders,the wind flow,etc.Among them,the flow of invaders can be divided into gravity flow,reverse flow,and jacking flow.In addition,this article also analyzes the lake shores,islands in the lake,shallow water areas,the effect of sand mining activities and ship navigation on the lake currents,analyzes the image features of different lakes,and obtains the overall characteristics of the Poyang Lake.Secondly,When the surface water of the lake carries sediment or other materials,it often leaves the linear texture information of the flow marks in the remote sensing image,and the texture information in the front of the water flow is more clear and stable,and the linear texture information in the back direction of the water flow is not continuous and clear.Therefore,by analyzing and studying this kind of texture information,the flow state information of the lake current can be obtained ?Through the analysis of remote sensing images with different resolutions at the same time,the texture features of different types of lake currents in optical images are summed up.Finally,a 30 m resolution remote sensing image is selected as the data source for comparison and analysis of water information in images.After analyzing the spectral characteristics of the water body information in the image,from the perspective of the spatial domain of the image,the texture of the image is analyzed using a gray-level cooccurrence matrix,and the effect of different scales on the texture extraction results is studied.From the frequency,by using radial scanning method,the radial distribution and angular distribution of texture information in different water bodies are analyzed to enhance the understanding of image texture information and provide the theoretical basis for the design of filter and algorithm.Aiming at the texture information of water body in medium resolution image,three algorithms for extracting the direction field of water body fluid flow are proposed.Algorithm one: Based on the direction field extraction algorithm,the algorithm Second: directional field extraction algorithm and algorithm based on frequency domain analysis Three: directional field extraction algorithm based on Gabor filter enhancement.The results show that the algorithm three can accurately reflect the directional field information of the study area with RMSE 7.11°,but the computation is complicated and the accurate response to the local information can not be given.The accuracy of the second algorithm is slightly worse than the third algorithm,RMSE is 12.74°,but it can reflect the details of the image,computing faster;The accuracy of algorithm one is the worst,RMSE is 14.99°,but it's the fastest.The experimental results show that the method based on image texture information proposed in this paper can quickly analyze the lake flow field.Therefore,it has certain theoretical and practical value for the study of the transfer situation of lake suspended solids.
Keywords/Search Tags:remote sensing, Poyang Lake, lake flow, Gabor filter, Huff transform
PDF Full Text Request
Related items