OpenCV - 如何从 uint8_t 指针创建 Mat

我有以下 C++ 代码：

void foo(const uint8_t* data, int height, int width) {
  // need to create a cv::Mat from *data, which is a pointer to grayscale image data

  // doesn't work correctly (compiles, but array access on the mat leads to a segmentation fault)
  auto img = cv::Mat(height, width, CV_8UC1, &data);

  // how can I fix the line above to create a proper cv::Mat?
}

// I'm calling foo like this
// img is a grayscale image
foo(img.ptr<uint8_t>(0), img.cols, img.rows);

谁能指出我在 foo 内部创建矩阵的语法有什么问题吗？

在本页cv::Mat 类参考，我们可以找到cv::Mat构造函数如下：

///! 2017.10.05 09:31:00 CST
/// cv::Mat public construction

Mat ()
Mat (int rows, int cols, int type)
Mat (Size size, int type)
Mat (int rows, int cols, int type, const Scalar &s)
Mat (Size size, int type, const Scalar &s)
Mat (int ndims, const int *sizes, int type)
Mat (int ndims, const int *sizes, int type, const Scalar &s)
Mat (const Mat &m)
Mat (int rows, int cols, int type, void *data, size_t step=AUTO_STEP)
Mat (Size size, int type, void *data, size_t step=AUTO_STEP)
Mat (int ndims, const int *sizes, int type, void *data, const size_t *steps=0)
Mat (const Mat &m, const Range &rowRange, const Range &colRange=Range::all())
Mat (const Mat &m, const Rect &roi)
Mat (const Mat &m, const Range *ranges)

创造cv::Mat from uint8_t pointer，我们可以使用这两个函数：

Mat (int rows, int cols, int type, void *data, size_t step=AUTO_STEP)
Mat (Size size, int type, void *data, size_t step=AUTO_STEP)

这是我的实验：

///! 2017.10.05 09:40:33 CST
/// convert uint8_t array/pointer to cv::Mat

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

int main(){        
    uint8_t uarr[] = {1,2,3,4,5,6,7,8,9,10,11,12};
    int rows = 2;
    int cols = 2;
    cv::Size sz(cols,rows);

    cv::Mat mat1(sz,CV_8UC3, uarr);
    cv::Mat mat2(rows, cols, CV_8UC3, uarr);

    std::cout<< "mat1: \n"<<mat1 << "\n\nmat2:\n" << mat2 << std::endl;
    return 0;
}

预期结果：

mat1: 
[  1,   2,   3,   4,   5,   6;
   7,   8,   9,  10,  11,  12]

mat2:
[  1,   2,   3,   4,   5,   6;
   7,   8,   9,  10,  11,  12]