利用主动轮廓和MRF探测震后损坏建筑物

发布时间：2018-04-20 21:27

本文选题：地震 + 损坏建筑物探测　；参考：《中国矿业大学学报》2014年01期

【摘要】：针对仅利用遥感影像不能有效探测震后损坏建筑物及其数目的问题,提出一种利用改进主动轮廓算法和马尔可夫随机场模型(CVMRF)探测震后损坏建筑物方法.利用震前GIS数据在震后遥感影像上确定震前建筑物区域,并使用主动轮廓模型提取其同质区;计算所提取的同质区与震前建筑物形状的相似度,并利用双阈值法获取初步的震后损坏建筑物探测结果;然后基于引入距离权重的马尔可夫随机场建立建筑物之间的空间关系模型,依此对那些难确定是否损坏的建筑物作进一步处理.实验结果表明:所提出的方法能有效地提高利用遥感影像探测损坏建筑物的精度与可靠性,为震后损坏建筑物探测及紧急救援提供有效技术手段.
[Abstract]:In order to solve the problem that only remote sensing images can not effectively detect damaged buildings after earthquake and their number, an improved active contour algorithm and Markov random field model (CVMRF) are proposed to detect damaged buildings after earthquake. Using pre-earthquake GIS data to determine the pre-earthquake building area on the remote sensing image after the earthquake, and using active contour model to extract the homogeneous area, and to calculate the similarity between the extracted homogeneous area and the pre-earthquake building shape. The preliminary detection results of damaged buildings after earthquake are obtained by using double threshold method, and then the spatial relationship model of buildings is established based on the Markov random field with distance weight. Further treatment of buildings that are difficult to determine whether or not damage is made. The experimental results show that the proposed method can effectively improve the accuracy and reliability of detecting damaged buildings using remote sensing images and provide an effective technical means for the detection and emergency rescue of damaged buildings after earthquake.
【作者单位】：中国矿业大学国土环境与灾害监测国家测绘局重点实验室环境与测绘学院;香港理工大学土地测量与地理资讯系;
【基金】：国家自然科学基金项目(41201451,41272389) 江苏高校优势学科建设工程项目中央高校基本科研业务费专项资金项目(2012LWB31) 江苏省普通高校研究生科研创新计划项目(CXZZ12_0941)
【分类号】：TP751;TU746.2

【参考文献】