一种基于最小错分概率的遥感影像变化检测算法

变化检测是研究自动实现目标变化信息提取的方法。为适应不同时期遥感影像成像环境和灰度的差异,提出一种基于最小错分概率的遥感影像变化检测方法,对待检测的遥感影像利用差影法提取出差值影像,利用最小错分概率的计算方式计算变化阈值,提取出重要的变化区域,最后采用去噪扫描和形态学的方式剔除噪声带来的虚假变化。实验结果表明,所提出的变化检测算法稳健、实时性强,具有较大的实用价值。

A Change Detection Algorithm Based on Minimal Classifying Error

Change detection is an approach for finding the target''s changing information automatically.In order to suit for the circumstance and gray differences of the remote sensing images of the same scene taken at different times,a change detection algorithm is presented to remote sensing images based on the threshold of minimal classifying error.The algorithm constructs difference image by subtraction operation,then,processes difference image by segmentation,with the threshold calculated by the principle of minimal classifying error,and finally,extracts the changed region after eliminates the false small changes caused by noises.Experiments show that the proposed method is robust and fast,which has preferable application value.