鉴于中值滤波算法中排序次数多、程序运行效率低的缺点,提出了一种快速的中值滤波算法.首先,利用统计理论排序法对滤波窗口的像素全面排序比较;在此基础上,利用相邻滤波窗口行列信息之间的相关性原理,通过迁出、移入一列新像素,同时保留剩余列像素的排序信息,从而快速查找到滤波窗口的中值;最后,通过仿真实验对比得出所提算法有效地降低了排序比较次数和算法的复杂度,满足图像处理的实时性.
A fast median filtering algorithm was proposed to improve complex sorting and low efficiency in the Median filtering algorithm.First of all,the pixels of filter window were sorted by the order statistics theory.Based on this, the median of filter window was quickly found by using the relationship between the adjacent filtering window category information.And a new pair of pixels was migrated,while the ordering information of the remaining column pixels was reserved.Finally,the results of contrast experiment were presented to verify the high efficiency of the proposed method.