Fuzzy comprehensive evaluation-based color image median filtering method and system

Abstract

The present invention discloses a fuzzy comprehensive evaluation-based color image median filtering method and system, wherein the method comprises the steps of determining a criterion set and a comment set, wherein the criterion set comprises the die lengths L of color vectors, an angle alpha between the color vectors, the chroma H, the saturability S and the intensity I, and the comment set comprises moderation, small deviation and large deviation; calculating the L, alpha, H, S and I values of all pixel point color vectors; constructing a fuzzy comprehensive evaluation matrix of a fuzzy comprehensive evaluation model according to the calculated values; weighting the fuzzy comprehensive evaluation matrix, and carrying out the weighting calculation to obtain the evaluation result vectors of all pixel point color vectors; then determining an optimal evaluation result vector; and finally moving a filtering window on a color image, repeating the above steps until all pixel points in the color image are traversed. By the embodiment of the present invention, the technical problems that how the original image color information and the noise information can be distinguished clearly, and how the salt-and-pepper noise pollution can be filtered strictly, are solved.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of image filtering, in particular to a color image median filtering method and system based on fuzzy comprehensive evaluation. Background technique [0002] Color image median filtering is to define a vector sequence structure for the color vectors in the color image, and then select the vector at the "middle" position as the median vector according to this sequence structure. For example, for a color image I with a size of m×n, in the direction of median filtering of a classic color image, it is first necessary to select a filter window with a size of (2k+1)×(2k+1), and then let this filter window be filtered Move pixel by pixel on the image, and every time it moves to a position, the pixel values ​​of the pixels in the window at this time are sorted through a certain sequence structure, and the (2k+1)th in the middle is selected 2 The / 2+1 vectors are determined as the new pixel ...

AI Technical Summary

Problems solved by technology

[0009] The main purpose of the embodiments of the present invention is to provide a color image median filtering method based on fuzzy comprehensive evaluation, which at least partially solves how to ensure the visual quality of the image during the filtering process, and can clearly distinguish the color information of the original image from the original image. Noise information, and can strictly filter out the technical problems of salt and pepper noise pollution

Images