论文部分内容阅读
通常情况下,受脉冲噪声污染的图像采用中值滤波法复原,受高斯噪声污染的图像采用均值滤波法复原。为了去除两者的混合噪声,Lee和Kassam提出了一种改进的均值滤波算法ModifiedTrimmedMean(MTM),但是MTM的使用受到了阈值的限制。为了在滤除退化图像中混合噪声的同时能更好地保护图像细节,我们详细分析了MTM滤波的特点,在对MTM进行改进的同时,提出了一种改进的自适应中值滤波算法(统计滤波)。该算法无需噪声的先验知识,利用VisualC++自动搜索阈值来实现图像的最佳复原。利用两种客观标准进行评价,实践证明新方法的处理结果优于传统的MTM方法。
Normally, the image contaminated by impulsive noise is reconstructed by means of median filtering, and the image contaminated by Gaussian noise is restored by mean filter. In order to remove the mixed noise, Lee and Kassam proposed an improved MeanTrimmedMean (MTM) algorithm, but the use of MTM is limited by the threshold. In order to better protect the image detail while filtering the degraded image, we analyze the characteristics of MTM filter in detail. At the same time we improve the MTM and propose an improved adaptive median filtering algorithm (statistics Filtering). The algorithm without prior knowledge of noise, the use of Visual C + + automatic search threshold to achieve the best image restoration. Using two objective criteria for evaluation, the practice proved that the new method is better than the traditional MTM method.