中文摘要：

提出一种改进的遥感影像自动噪声评价方法,基本思想是通过自动提取最优平坦区域来检测整幅影像的噪声水平。首先使用低通滤波器过滤原始影像以初步估计影像的噪声水平,然后根据初步估计结果,选用最优的滤波器组合提取均匀平坦区域,再对均匀平坦区域自适应迭代优化,最后对多种噪声评价指标进行优化组合,实现对影像噪声水平的评价。

英文摘要：

We propose a modified automatic noise evaluation method of remote sensing image.The basic idea is to evaluate the noise level of remote sensing image based on the optimum flat sites which are picked automatically.Firstly,using low pass filter to filtrate the original image and estimate the noise level preliminary.Secondly,pick up the flat sites on the basis of the optimum filter combination and the result of last step.Then adaptively and iteratively optimize those flat sites.Lastly,use the optimized combination of the noise evaluation indices to evaluate the noise level of the remote sensing image.