图像处理:亮度加权

Image processing: luminance weighted

我想权衡亮度值

示例: 我有一个亮度值向量:

vector <int> lum {50,100,150,200,250);

我有一个系数向量:

vector <float> coef {5.1 , 2.55 , 1.7 , 1.275, 1.02 }

我创建了一个空图像:

Mat1 m(15):

所以,我的第一个亮度值是 50 (lum[0]=50),我希望它应用于矩阵的 5.1 (coef[0]=5.1) 第一个像素。为此,我需要用第一个和第二个亮度值对第 6 个像素进行加权。在我的例子中,我的第 6 个像素的亮度将是 95,因为 (0.1*50)+ (0.9*100)=95

目前,对于第二个系数(coef[1]=2.55),我在之前的计算中使用了 0.9 on 2.55。该系数仍为 1,65,因此第 7 个像素的亮度为 100,第 8 个像素的亮度为 (0.65*100)+ (0.35*150) = 117,5.

等等...

其实我有这个:

//Blibliothèque Opencv
#include "opencv2/core/core.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/imgcodecs.hpp"
#include "opencv2/highgui/highgui.hpp"
// cpp include
#include <iostream>
#include <cmath>
#include <math.h>
#include <string.h>
#include <vector>
#include <opencv2/opencv.hpp>

#define MPI 3.14159265358979323846264338327950288419716939937510
#define RAD2DEG (180./MPI)

using namespace cv;
using namespace std;


vector <int> lum{ 50, 100, 150, 200, 250 };
vector <float> coef (5,0);
vector <int> newv(15, 0);
float pixelRef = 255, angle = 0, coeff = 0;


int main()
{


    for (int n = 0; n < lum.size(); ++n)
    {
        //get angle
        angle = ((acos(lum[n] / pixelRef)));
        cout << "angle :" << angle*RAD2DEG << endl;
        // get coefficient
        coef [n] = (1 / (cos(angle)));
        cout << "coeff :" << coef [n] << endl;
        // try to weighted my pixels
        newv[n] = (coef*lum[n]) + ((1 - coeff)*lum[n + 1]);
    }
    return 0;
}

我将 coef 的最后一个元素修改为 3.02f 以表明此代码也能很好地处理最后一个元素。结果序列为:

50, 50, 50, 50, 50, 95, 100, 117.5, 150, 182.5, 218.75, 250, 250,

代码可能 re-written 更好,但我会把它留给你:

#include <opencv2/opencv.hpp>
#include <vector>
#include <iostream>
using namespace cv;
using namespace std;

int main()
{
    vector <int> lum{ 50, 100, 150, 200, 250 };
    vector <float> coef{ 5.1f, 2.55f, 1.7f, 1.275f, 3.02f };

    vector<float> v;

    int idx_lum = 0;
    int idx_coef = 0;

    while (true)
    {
        int c = int(coef[idx_coef]);
        for (int i = 0; i < c; ++i)
        {
            v.push_back(float(lum[idx_lum]));
        }

        float alpha = fmod(coef[idx_coef], 1.f);
        float beta = 1.f - alpha;

        v.push_back(alpha * lum[idx_lum] + beta * lum[idx_lum + 1]);

        idx_lum++;
        idx_coef++;
        coef[idx_coef] = coef[idx_coef] - beta;

        if (idx_lum >= lum.size() - 1 || idx_coef >= coef.size() - 1)
        {
            int cc = int(coef[idx_coef]);
            for (int i = 0; i < cc; ++i)
            {
                v.push_back(float(lum[idx_lum]));
            }

            // Only if the last remainder is needed
            //float alpha = fmod(coef[idx_coef], 1.f);
            //v.push_back(alpha * lum[idx_lum]);
            break;
        }
    }

    // Print out the values
    copy(v.begin(), v.end(), ostream_iterator<float>(cout, ", "));

    // Get a cv::Mat from the std::vector
    Mat1f m = Mat1f(v).t();

    return 0;
}