Detection Technology Research Of Earthquake Change In Remote Sensing Image

At 14:28:04 on May 12th,2008, the magnitude-8.0 strong quake, happened in Wen Chuan County and Bei Chuan County, Sichuan Province, has been the most destructive and pervasive earthquake since the founding of New China. The earthquake caused great damage for the country and the people. Since the earthquake presents such features as activity of high frequency, great intensity and the comprehensive destruction, how to acquire the accurate and detailed seismic data quickly becomes the key of loss statistics, estimation of earthquake and reconstruction following the calamity.Seismic remote sensing image change detection is the registration of the same region have been before the earthquake and after the earthquake to compare two remote sensing images, seismic data to obtain the disaster area, while according to data on earthquake damage analysis and assessment is given a technical. Given the current ease of access to remote sensing images and remote sensing images covering a wide range of characteristics, the use of remote sensing image analysis of large-scale detection of changes in the regional information has become one of the hot, and has been used in many fields, such as land desertification research, pest prevention and detection urban green area of detection, and fire, drought, earthquake and other natural disasters in the detection and assessment.Currently, there are many about the application of remote sensing image change detection techniques, but the earthquake caused a wide range of earthquake zones, artificial force on the affected point is very difficult to confirm, in the mountains, valleys, forests are widely distributed in the affected areas even more so. Therefore, due to the uncertainty of the affected sites, the current change detection method can not determine which changes in the detection of earthquakes is more efficient and more accurate. Therefore, this paper start from a variety of detection technology transformation, according to various changes in comprehensive analysis of test results, on the earthquake disaster zone and the destruction gives a comprehensive analysis and evaluation.The paper mainly analyzes the following respects:The thesis first introduces the research background and meaning of change detection technology in remote sensing image, and summarizes the latest progresses development and the research trend nowadays.Secondly, aim at the research status and existing problems of remote sensing image change detection in earthquake, this dissertation offers theoretical elaboration and comprehensive analysis of change detection technologies which are used in the paper, mainly including otsu threshold, the segmentation for target region in gray-scale images, the two-way change detection, the normalized difference vegetation index (NDVI), vegetation coverage, grey-scale morphology, operators of morphological edge detection, and so on.Thirdly, in this essay otsu threshold of the two-way changes, vegetation coverage to detect changes and the edge of the extraction and the image of the value of integrating all three types of changes in microsoft visual inspection techniques and various operating system platforms 6.0 finish the design and implementation.Fourthly, several critical parameters of the three change detection technologies are presented, and the information of earthquake disaster is evaluated from two aspects of the changes in surface features and the destruction of vegetation, according to many testing results.At last, a summation is made to generalize the work in this paper. Suggestions are proposed for further research and improvement.The innovations in the dissertation as bellow:Rapid Diagnosis of Remote Sensing Image Earthquake Change is developed under the environment of Microsoft Visual C++6.0. The image in the experiment is segmented 1/30 of the original remote sensing image. Large area of earthquake (many images) need to quickly change detection method to achieve rapid response.Algorithm is not complex, certainly much faster than the MATLAB platform, much faster than the complex algorithm, such as wavelet transform.