测绘学报 ›› 2014, Vol. 43 ›› Issue (9): 945-953.doi: 10.13485/j.cnki.11-2089.2014.0138

• 学术论文 • 上一篇    下一篇

融合多特征的遥感影像变化检测方法

李亮1,2,舒宁3,王凯4,龚龑4   

  1. 1. 四川省第三测绘工程院
    2. 武汉大学 遥感信息工程学院
    3. 武汉大学遥感信息工程学院
    4. 武汉大学
  • 收稿日期:2013-12-12 修回日期:2014-01-08 出版日期:2014-09-20 发布日期:2014-09-25
  • 通讯作者: 李亮 E-mail:liliang1987wuda@163.com
  • 基金资助:

    国家自然科学基金;中央高校基本科研业务费专项基金资助;中央高校基本科研业务费专项基金资助

Change Detection Experimental Study Based on Spectral and Texture Features

LI Liang1,2,SHU Ning1,WANG Kai1,GONG Yan1   

  1. 1. School of Remote Sensing and Information Engineering, Wuhan University
    2. The Third Academy of Engineering of Surveying and Mapping
  • Received:2013-12-12 Revised:2014-01-08 Online:2014-09-20 Published:2014-09-25
  • Contact: LI Liang E-mail:liliang1987wuda@163.com

摘要:

本文提出了一种面向对象的多特征融合的变化检测方法。首先通过影像分割获取像斑,然后统计各像斑的光谱直方图和LBP(local binary patterns)纹理直方图,利用G统计量计算不同时期像斑之间的光谱距离和纹理距离,采用自适应的方法将光谱距离和纹理距离加权构建像斑的异质性,最后结合EM(expectation maximization)算法和贝叶斯最小错误率理论获取像斑的变化类别。在QuickBird影像上的实验表明该方法能够充分融合光谱特征和纹理特征,从而提高变化检测的精度。

关键词: 面向对象, 多特征融合, LBP纹理, G统计量, EM算法

Abstract:

In order to make full use of spectral and texture features, an object-oriented change detection method for remote sensing images based on multi-features fusion is proposed in this paper. First image segmentation is used to get image objects. Then the spectral and lbp texture histograms of each object are extracted. G statistic is adopted to calculate the distance of histograms between two periods. The heterogeneity of each object is built by weighted spectral and texture distance. At last, the expectation maximization algorithm and bayesian rule with minimum error rate are applied to get the change/no change results. Experimental results on QuickBird and SPOT-5 images show that the method proposed in this article can integrate the spectral and texture features effectively and improves the accuracy of change detection.

Key words: object-oriented, multi-features fusion, local binary patterns, G statistic, expectation maximization

中图分类号: