一种改进的自适应中值滤波方法

摘要

The median filtering, one nonlinear filtering method is widely used owing to its effective noise suppression. An improved adaptive median filtering method is proposed in this study because that the traditional adaptive median filtering easily divides the high frequency signal points into noise points, and make image details blurred. A minimum set distance measure is done while the noise points is detecting. Results show that the improved method is better than origin algorithm in detecting accuracy, noise reducing and keeping details.%非线性滤波方法中中值滤波因其有效的噪声抑制技术,得到广泛应用.针对传统的自适应中值滤波易将高频信号点误分为噪声点,同时容易模糊图像细节的问题,提出一种改进的自适应中值滤波方法.在进行噪声点检测时,引入了最小集合距离测度,有效地避免了将高频信号误判为噪声.实验结果表明,该方法在检测正确率、降噪和保留细节方面都优于改进前的算法.