一种高效的基于阈值的图像滤波算法及其实现

An Efficient Threshold Based Image Filtering Algorithm and Its realization

通过对均值滤波和中值滤波进行分析,提出并实现了一种新的滤波算法。该算法通过设置一阈值,当像素点的值和邻域均值的差值大于该阈值时,才取邻域均值为该点的值,否则该点的值保持不变。那么图像就不会模糊又去除了噪音,也不用排序,就能取得和中值滤波一样的滤波效果,速度比中值滤波快得多。而且该算法能够克服均值滤波和中值滤波的不足,在均值滤波和中值滤波无法胜任的领域仍能大显身手,展现出良好的适应性和高效性。并可根据需要对滤波过程实现控制,具有很大的灵活性。

<Abstrcat>A new image filtering algorithm is put forward and realized by analyzing the mean value filter and the medium value filter. A threshold is set to compare with the mean value of the pixel. If the difference between the two values is bigger than the threshold, the pixel gets value of the mean value, otherwise its value keeps invariable. So the image keeps clear and noise is got rid off. It needn't sorting, so the filtering speed is much faster than that of the medium value filter, at the same time the filtering result is as good as the medium value filter. This algorithm can overcome the deficiency both in the mean value filter and the medium value filter. It is competent anywhere the mean value filter and the medium value filter can't work. It also shows strong adaptability and high efficiency. The filtering process can also be controlled flexibly to meet our need.