Change Detection Method For Remote Sensing Images Based On Multi-features Fusion

The technology of remote sensing image to earth observation has the advantages of covering area,periodicity and wide information.Based on the change detection of remote sensing image,it can continuously observe the change of surface features,and can comparatively analyze the change information and the change process quantitatively.Using change detection technology can quickly obtain the results of urban expansion,natural disaster assessment,geographical conditions monitoring and battlefield situation analysis,which make the study on high resolution remote sensing image change detection become one of the hotspots in the field of remote sensing.The existing high resolution remote sensing image change detection algorithm(single feature method)has advantages and disadvantages as well as the scope of application.Combining with a variety of features can effectively use the advantages of each feature to improve the detection accuracy of the algorithm,which provides new ideas for change detection.Aiming at the characteristics of high resolution remote sensing image,this paper makes a systematic research on object-oriented change detection technology.This paper also puts forward the corresponding improvement methods to solve the problems of existing methods.The main contents are as follows:(1)On the basis of summarizing and analyzing the current study at home and abroad and existing problems of the high resolution remote sensing image change detection,this paper introduces the technical route of change detection in detail.According to the different processing units,the change detection of high resolution remote sensing image is divided into pixel based and object oriented change detection,and the two methods are discussed detailedly.(2)In order to solve the problem of high matching error rate of the remote sensing image in the similar scene,the SIFT feature matching algorithm combined with global information is proposed: in this paper,we propose a new method of principal direction assignment for DAISY descriptor.The transformed DAISY descriptor is invariant to rotation;A global texture information descriptor is introduced into the DAISY descriptor.When the image is locally similar,the global texture information is used to describe it,which reduces the probability of error matching due to the local region similarity.The experimental results show that the algorithm has strong robustness to the image transformation of small angle rotation and large spectral change etc.(3)The traditional-oriented change detection algorithm mainly depends on the spectral information of the speckle image,it can not effectively utilize the advantage of the multi feature detection.Based on the object-oriented analysis,a multi-feature fusion algorithm for remote sensing image change detection is proposed.Firstly,the color histogram and edge straight line gradient histogram of each object are calculated based on multiscale segmentation of image objects;Then,the EMD operator is used to estimate the color distance and the distance between the edge lines of the objects in different periods.The distance between the color feature distance and the edge line distance is combined to construct the heterogeneity of the object by using the method of sel-adaptive weighting;Finally,the change detection results of the image are obtained by using histogram curvature analysis.Experimental results show that the proposed method can fully fuse the color and edge line features and improve the accuracy of the change detection.