针对传统中值滤波算法比较运算量大、处理效率低、无法满足实时性的问题,以中值滤波原理为基础,对滤波排序算法和实现方案进行了研究,提出了一种基于System Generator的快速中值滤波算法.该算法通过两次行、列排序将3×3窗口中9个像素取中值简化为3个像素的排序运算,使得单窗口查找中值的比较次数由传统排序算法的36次最少降到了14次.结合System Generator系统建模工具,将处理速度提升到传统方法的近六倍,达到了快速抑制噪声的目的,满足了图像实时处理的要求.%Aiming at traditional median filtering algorithm with high time complexity, low processing efficiency and poor real-time, this paper researched on filtering sorting algorithm and its implementation based on the principle of the median filter,and proposed a fast median filtering algorithm based on the generator.The algorithm took the 3×3 window of nine pixels sorting operation into a three-pixel sorting operation after rows sorting and columns sorting, it reduced the number of comparisons from 36 times of the traditional sorting algorithm to a minimum of 14 times in a single window.Combined with System Generator, the algorithm enhances the processing speed nearly 6 times as high as the conventional method, which can achieve the purpose of fast noise suppression and meet the requirements of real-time image processing.
展开▼