文章提出一种利用局部统计量消除图象中与信号强度有关的噪声的算法。用二段直线近似代表一个局部窗口中各象元灰度值的概率分布,因而使各局部统计量的计算大为简化,由于各象元可分別处理,因而本方法易于实时处理及并行运算。实验表明所提方法对较宽范围的噪声比现有的用局部统计量的算法效果有很大改进。 This paper presents an algorithm that uses local statistics to eliminate signal-dependent noise in the image. Using two straight line approximations to represent the probability distribution of the gray value of each pixel in a local window, the calculation of each local statistics is greatly simplified. Because each pixel can be processed separately, the method is easy to process in real time and parallel. Experiments show that the proposed method can greatly improve the performance of a wide range of noise than the existing local statistics.