摘要
针对含有椒盐噪声和高斯噪声的灰度图像,研究一种新型的自适应混合滤波算法。首先利用改进的自适应中值滤波算法过滤图像中的椒盐噪声;其次利用改进的自适应均值滤波算法过滤图像中的高斯噪声。这种混合滤波算法具有自适应扩大滤波窗口以及自适应选择最佳阈值的特点。计算机仿真实验证实,该方法不仅在改善信噪比(SNR)和最小均方误差(MSE)上明显优于传统的中值滤波、均值滤波、小波硬阈值、软阈值等方法,而且能较好地保护图像细节。
In this paper,a new adaptive mixed filter is investigated for noise removal of an image contaminated by Guass noise and impulse noise. In the process of removing pepper noise, this method can enlarge filter window size adaptively. While in the process of removing Gauss noise, it can choose the threshold adaptively. Computer simulation results validate that the denoising ability of this method is better than the traditional mean filtering method, median filtering method, wavelet hard threshold method and wavelet soft threshold method by its higher Signal to Noise Rate （SNR） and smaller minimizes the Mean Squared Error（MSE）. The eminent performance of this method is also proved that it has better details preservation capacity.
出处
《现代电子技术》
2009年第10期96-99,共4页
Modern Electronics Technique
