Research On Building Damage Information Extraction Technology Of High Resolution Remote Sensing Image

Buildings in the city are an important place for people to live in. When the destructive earthquake happen and damage to buildings are often very serious, there is a great impact to people’s life and property safety, so there is an urgent need to solve the problem including fast and accurate information extraction of building damage after earthquake, and reliable basis for emergency rescue decision. With the development of space science and remote sensing technology, high resolution remote sensing image is more and more widely used. High resolution remote sensing images can provide more detailed surface information, such as spectral information, geometric information, texture information and context information, Object oriented remote sensing image analysis technology provides a new idea for remote sensing image information extraction, and it also provides a greater possibility for the accurate extraction of seismic damage information based on the above characteristics of high resolution remote sensing image.Object oriented remote sensing image analysis method has been widely used in many fields, but it is not very mature in the seismic damage information extraction of remote sensing image building, In this paper, the seismic damage information of buildings in high resolution remote sensing image is studied by using the remote sensing image of the former Ludian earthquake in August 3, 2014. Firstly, the object oriented image segmentation is carried out for the high resolution remote sensing image of the two phase, This paper proposes improved multi-scale segmentation method, the main idea is to detect the edge of remote sensing image after preprocessing and get edge information of the buildings, On the basis of this, the heterogeneity of the minimum region merging algorithm is used for image segmentation. On the basis of using heterogeneous minimum region merging algorithm for image segmentation so as to avoid the over segmentation phenomenon caused by the different spectral information appears in the same building. Then the object oriented fuzzy classification technology is used to extract the building information of the two time phase. At the same time, the change information of the building in the two time phase is detected by the step change detection method with multi features::Firstly, image change area is detected based on pixel level, Then, the change area is detected based on the geometric feature and texture feature, and the information of building change is extracted. Finally, the GIS spatial overlay analysis method was used to obtain the damaged building area, and the damage degree of buildings was evaluated according to the correlation coefficient.In this paper, the date come from Ludian Yunnan earthquake, the building damage information was extracted by the above method. The experimental results show that the improved method proposed in this paper can improve the accuracy of seismic damage information extraction.