Research On The Extraction Of Water Body From Envisat Remote Sensing Image

As the carrier of water resource in city,urban rivers have important functions in landscape ecology,flood control and Mitigating heat island effect.Thus,it is significant to recognize and extract the urban river.Because of the spatial resolution of remotely sensed imagery becoming higher,the traditional technology of classification based on pixels for low and medium resolution remotely sensed imagery can't satisfy the requirement in the processing of remotely sensed imagery,and is instead of the object-oriented method of features recognition and extracting gradually Using object-oriented algorithm,land feature in city,such as road,building,river,green,land,can be extracted readily.In remote sensing images,water is an important geographical target.Effectively extracting water information is of great significance for water conservancy planning and ship navigation.Remote sensing imaging has a short cycle time,strong real-time performance,high resolution,and low technical threshold.It has gradually become an important means for extracting water information.With the development of digital image processing technology,it has become possible to automatically extract water target information from remote sensing images.At present,there are a large number of remote sensing water extraction methods,mostly for multi-band remote sensing color images.This article is aimed at radar remote sensing images,which are grayscale images.Water extraction methods for such remote sensing images include the Fisher criterion-based threshold separation algorithm,edge-based segmentation method,and Graph Cuts-based method.Aiming at the problem of self-extraction of water in microwave remote sensing image,combined with the distribution of gray value of the image,the assumption that the gray value of water body and non-water body obey the Gaussian distribution is made.A Gaussian mixture model is proposed to propose an image segmentation based on Graph Cuts.algorithm.According to the distribution of pixel gray value,make an assumption that water and non-water gray value follow Gaussian distribution respectively,and present an algorithm of image segmentation based on GMM to achieve the purpose of self-extraction of water bodies in microwave remote sensing images.First,the real remote sensing images which are processed by Gauss filter are tested.Then,the EM algorithm is used to calculate the relevant parameters of the Gaussian Mixture Model and the model is used to calculate the weight of t-link in Graph Cuts.The augmented path method is used to solve the maximum flow minimum cut problem for image segmentation.In this paper,compared with the correlation algorithm of threshold separation,the Graph Cuts algorithm which combines the Gaussian Mixture Model achieved higher accuracy with the manual annotation released.