基于快速EM算法和模糊融合的多波段遥感影像变化检测

CHANGE DETECTION METHOD OF MULTIBAND REMOTE SENSING IMAGES BASED ON FAST EXPECTATION-MAXIMIZATION ALGORITHM AND FUZZY FUSION

提出了一种基于快速EM(expectation maximization)算法和模糊融合的多波段遥感影像无监督变化检测方法.该方法首先对各波段差异影像采用基于直方图分析的快速EM迭代算法获取变化分类阈值和变化信息,随后对各波段的变化信息进行模糊融合和判决,生成最终的变化检测图.利用真实的多波段遥感影像进行了实验,本文方法在运行时间和检测效果两个方面都具有优越性.

An unsupervised change detection method based on fast expectation-maximization（EM） algorithm and fuzzy fusion for multi-band remote sensing images was proposed.First,fast EM iteration algorithm based on histogram of image difference in each band was used to obtain the change class threshold and change information.Second,the fuzzy theory and relationship matrix were adopted to integrate the classification information of all bands,and the final changed and un-changed map of the bitemporal remote sensing images were obtained.Thus,the change detection image was formed.The real bitemporal SPOT5 and Landsat TM satellite imagery were performed to evaluate the effectiveness of the proposed method.The results show that the proposed method reduces the processing time and gets better detection effectiveness comparing with other methods.