OpenCv 027---均值迁移滤波

Posted 2020-11-23 vince-wu

1 前备知识

    均值迁移模糊是图像边缘保留滤波算法中的一种，经常用在对图像进行分水岭分割之前去噪声，可以大幅度提升分水岭分割的效果。均值迁移模糊的主要思想如下： 就是在图像进行开窗的时候，考虑像素值空间范围分布，只有符合分布的像素点才参与计算，计算得到像素均值与空间位置均值，使用新的均值位置作为窗口中心位置继续基于给定像素值空间分布计算均值与均值位置，如此不断迁移中心位置直到不再变化位置（dx=dy=0），但是在实际情况中我们会人为设置一个停止条件比如迁移几次，这样就可以把最后的RGB均值赋值给中心位置。

2 所用到的主要OpenCv API

/** @brief Performs initial step of meanshift segmentation of an image.
The function implements the filtering stage of meanshift segmentation, that is, the output of the
function is the filtered "posterized" image with color gradients and fine-grain texture flattened.
At every pixel (X,Y) of the input image (or down-sized input image, see below) the function executes
meanshift iterations, that is, the pixel (X,Y) neighborhood in the joint space-color hyperspace is
considered:
f[(x,y): X- 	exttt{sp} le x  le X+ 	exttt{sp} , Y- 	exttt{sp} le y  le Y+ 	exttt{sp} , ||(R,G,B)-(r,g,b)||   le 	exttt{sr}f]
where (R,G,B) and (r,g,b) are the vectors of color components at (X,Y) and (x,y), respectively
(though, the algorithm does not depend on the color space used, so any 3-component color space can
be used instead). Over the neighborhood the average spatial value (X‘,Y‘) and average color vector
(R‘,G‘,B‘) are found and they act as the neighborhood center on the next iteration:
f[(X,Y)~(X‘,Y‘), (R,G,B)~(R‘,G‘,B‘).f]
After the iterations over, the color components of the initial pixel (that is, the pixel from where
the iterations started) are set to the final value (average color at the last iteration):
f[I(X,Y) <- (R*,G*,B*)f]
When maxLevel > 0, the gaussian pyramid of maxLevel+1 levels is built, and the above procedure is
run on the smallest layer first. After that, the results are propagated to the larger layer and the
iterations are run again only on those pixels where the layer colors differ by more than sr from the
lower-resolution layer of the pyramid. That makes boundaries of color regions sharper. Note that the
results will be actually different from the ones obtained by running the meanshift procedure on the
whole original image (i.e. when maxLevel==0).
@param src The source 8-bit, 3-channel image.
@param dst The destination image of the same format and the same size as the source.
@param sp The spatial window radius.
@param sr The color window radius.
@param maxLevel Maximum level of the pyramid for the segmentation.
@param termcrit Termination criteria: when to stop meanshift iterations.
 */
CV_EXPORTS_W void pyrMeanShiftFiltering( InputArray src, OutputArray dst,
                                         double sp, double sr, int maxLevel = 1,
                                         TermCriteria termcrit=TermCriteria(TermCriteria::MAX_ITER+TermCriteria::EPS,5,1) );

3 程序代码

#include <opencv2/opencv.hpp>
#include <iostream>

using namespace cv;
using namespace std;

int main(int artc, char** argv) {
    Mat src = imread("images/Demo.jpg");
    if (src.empty()) {
        printf("could not load image...
");
        return -1;
    }
    namedWindow("input", CV_WINDOW_AUTOSIZE);
    imshow("input", src);

    Mat dst;
    pyrMeanShiftFiltering(src, dst, 15, 50, 1, TermCriteria(TermCriteria::COUNT + TermCriteria::EPS, 5, 1));
    imshow("result", dst);

    waitKey(0);
    return 0;
}

4 运行结果

略

5 扩展及注意事项

none

6*目前只做大概了解，知道有这一算法，后续具体使用再做具体分析

后面发现对着知识星球上的过程看倒不如直接看书.....  

所以现在写这些东西只是把它当成一种养成，只做很简单的了解，贴下API，浅入浅出，让自己知道有这么一回事....

回归课本，按照课本内容来会更高效一些，而且知识点也较全，看完课本再回过头将一些知识点往对应的博文上添加