Computer Engineering and Applications ›› 2012, Vol. 48 ›› Issue (3): 24-28.

• 博士论坛 • Previous Articles     Next Articles

Sensitive feature vector-based disaster change detection using SAR images

LI Kun1, YANG Ran1, WANG Leiguang2, LIN Liyu1, QIN Qianqing1   

  1. 1.State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China
    2.School of Environmental Science and Engineering, Southwest Forestry University, Kunming 650224, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-01-21 Published:2012-01-21


李 坤1,杨 然1,王雷光2,林立宇1,秦前清1   

  1. 1.武汉大学 测绘遥感信息工程国家重点实验室,武汉 430079
    2.西南林业大学 资环学院,昆明 650224


In order to overcome the problem of the limited application of optical image during bad weather condition, this paper analyzes the sensitivity of SAR image gray features and texture features for the four land cover changes(lake, landslide debris flow, part of collapsed buildings and serious collapsed buildings) which usually happen in the natural disasters, and put forward the concept of sensitive feature vector. Taking sensitive feature vector which is comprehensive utilization of the gray scale difference and texture difference as evaluation factor, a new disaster change detection method of SAR image is carried out, and verifying on two groups of ALOS SAR images prove the better performance than the method only using gray scale difference.

Key words: Synthetic Aperture Radar(SAR), texture difference, gray difference, change detection, disaster assessment

摘要: 为了解决阴雨云雾条件下光学遥感图像的应用局限性问题,针对典型的四类地表变化(堰塞湖、滑坡泥石流、部分倒塌建筑和严重倒塌建筑)分析SAR图像灰度和纹理特征的敏感程度,并提出敏感特征向量的概念;以综合利用了灰度差值和纹理差值的敏感特征向量作为评判因子,结合主成分分析技术和K均值聚类技术,提出了新的SAR图像灾害变化检测算法。该方法算法简单,检测效果好,并用两组ALOS SAR实验数据进行了证实。

关键词: 合成孔径雷达, 纹理差值, 灰度差值, 变化检测, 灾害评估