在采用平均技术的数字图象噪声平滑方法中,噪声方差的削减和图象分辨率的损失通常会伴随而至,即噪声的抑制是以图象分辨率的损失为代价的。为了解决这个矛盾,本文研究并提出了一种新的数字图象噪声平滑方法——松弛式Sigma滤波.该方法不仅可以有效地削减噪声方差,同时以它良好的适应性能可以保持图象原有的分辨率,这是一般噪声平滑方法所不能比拟的。 In the averaging technique of digital image noise smoothing methods, noise variance reduction and image resolution loss are usually accompanied by noise suppression at the expense of image resolution loss. In order to solve this contradiction, this paper researches and proposes a new method of smoothing digital image noise-relaxation Sigma filter, which can not only effectively reduce the noise variance, but also keep the original image with its good adaptability Resolution, which is unmatched by normal noise smoothing methods.