在OpenCV与c++中，用我的自定义最大值和自定义最小值规范化0到1之间的矩阵

Question

我想用c++在OpenCV中用我的自定义最大值和自定义最小值规范化0到1之间的矩阵。这是我的代码:

#include <iostream>
#include <opencv2/core/core.hpp>
int main()
{
    cv::Mat src(cv::Size(2, 2), CV_32FC1, cv::Scalar::all(0));
    cv::Mat dst;
    src.at<float>(0,0) = -3;
    src.at<float>(1,1) = 3;
    std::cout << "src: " << src << 'n';
    cv::normalize(src, dst, 0, 1, cv::NORM_MINMAX, CV_32FC1); // Not I would like!
    // I'd like to normalize between 0 and 1 but my minimum is -10 and my maximum
    // is 10 and not between [-3, 3]!
    std::cout << "dst: " << dst << 'n';
}

Answer 1

#include <iostream>
#include <opencv2/core/core.hpp>
int main()
{
    double min = -10;
    double max = 10;
    cv::Mat src(cv::Size(2, 2), CV_32FC1, cv::Scalar::all(0));
    cv::Mat dst;
    src.at<float>(0,0) = -3;
    src.at<float>(1,1) = 3;
    std::cout << "src: " << src << 'n';
    dst = (src - min) / (max - min);
    std::cout << "dst: " << dst << 'n';
    return 0;
}

Answer 2

你可以这样做(使用lambda函数):

dst = src.clone();
std::for_each(dst.begin<float>(), dst.end<float>()
               , [](float& ele) { ele = (10+ele)/20; });