Research And Application Of Key Technologies For River And Lake Morphological Change Assessment Based On Big Data Platform

The shape of river or lake can reflect its health,which plays an important role in the protection and management of rivers and lakes and the function of social services in rivers and lakes.It is necessary to manage and control the shape of rivers and lakes.However,the shape control of rivers and lakes is still unable to meet the needs of informationization,and the degree of automation is mostly low.The shape control of rivers and lakes still lacks a systematic theoretical basis for analysis and evaluation.At the same time,the morphological data is mostly obtained by remote sensing images,which has the characteristics of many periods and large amount of data.Therefore,combining big data technology with systematic analysis and evaluation of its morphological changes can provide a new idea for the management and control of automated and intelligent rivers and lakes.In the big data environment,the paper aims to realize the control of automatic rivers and lakes,and explores the research and design of the river and lake morphological change assessment model in the big data environment,and establishes a set of quantitative index evaluation system based on morphology.To provide a theoretical basis for the automation of rivers and lakes shape control,and selected Axe Lake as the pilot research object,and analyzed the relevant results.The main research contents and results are as follows:(1)In this paper,a river and lake morphological change assessment model combining big data technology is proposed,which combines OpenCV,GDAL,MapReduce,Spark and other related technologies,and Spark,is built and implemented.Customly reading of images in GeoTIFF format in HDFS is realized,and conversion of various formats of remote sensing images in tthe process of analysis is done.Various analysis results can be output and saved in JSON format.(2)In this paper,based on the big data platform,the water body shape extraction is completed,and based on the secondary development of GDAL and OpenCV,the information extraction and processing module of remote sensing image in GeoTIFF format is completed.The automatic sample selection strategy is implemented in support vector machine which is used to do water classification,and a sliding scale frequency statistics algorithm is proposed,which can more accurately determine the target pixel range,thus supporting the automatic processing of batch remote sensing images and the extraction of water contours.(3)In this paper,by studying the quantitative expression of the morphological characteristics of the lake,the assessment of the extracted morphological changes of the water body is realized,which provides a theoretical basis for the control of rivers and lakes.The box-counting dimension calculation algorithm is implemented based on OpenCV.And a set of index system for comprehensive quantitative analysis of lake morphological changes is established.(4)In this paper,a comprehensive computing environment based on OpenCV,GDAL,MapReduce,and Spark is built.The axe lake was selected as the pilot research object,and the results of its morphological change assessment were analyzed.It was found that the shape of the axe lake has the characteristics of a typical narrow lake and is greatly influenced by human factors.By studying the trends of its morphological characteristics within one year,the red line threshold of the morphological change assessment model was evaluated.