openCV:从矩阵中绘制色彩图

时间:2014-11-17 06:10:54

标签: c++ opencv plot

我有一个50 x 50矩阵,具有从1到50的某些值。我需要使用OpenCV绘制相同的颜色映射。

我已经在MATLAB中完成了这项工作(在this answer的帮助下):

>> A = randi([10,60],100,100);
>> colormap('hot')
>> imagesc(A)
>> colorbar    

输出就像this

我想使用openCV做同样的事情。

我该如何处理?

我无法理解如何实现这样的“热”功能(仅见热门)

like this

此外,我还想对第一张图片中显示的颜色条感兴趣。


更新(我正在运行的最终代码)

不幸的是我仍然无法绘制正确的颜色图:这是MATLAB代码:

注意:A是10乘100矩阵。

A = [0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ;     ]

颜色表( '热');  于imagesc(A)   彩条

这是输出:

enter image description here

现在我创建了一个包含相同数据的文本文件,我的文本文件如下所示:

enter image description here

使用以下代码实现与MATLAB相同的结果:

   int main() 


{




ifstream Read("myfile.txt");

vector<int> nums;
while ( !Read.eof() ) {
    int n;
    Read >> n;
    nums.push_back(n);
}
// now make a Mat from the vector:
Mat mat(nums);

cout<<mat<<endl; //for testing 


Mat mat1d(nums);
Mat mat2d = mat1d.reshape(1, 10);
//
Mat image; //create an empty image. (you can leave it empty ;)
//// Apply the colormap, but on the 2d mat, not on the 2d vector, please:
applyColorMap( mat2d, image, COLORMAP_JET );
            // Show the result:
imshow("colormap", image);
waitKey(0);
return 0;
}

但是上面的输出是毫无意义的。

我需要匹配MATLAB和Opencv的结果。

1 个答案:

答案 0 :(得分:1)

你必须将1d矢量重塑为2d垫:

    //
    // before doing anything else, CLEAN UP YOUR DAMN TXTFILE. 
    // it should contain nothing than numbers and spaces.
    ifstream Read("m.txt");

    // be extra picky about the type here.
    // we are reading in a 8bit grayscale map.
    vector<uchar> nums;
    for (int i=0; (i<100*10)&&(!Read.eof()); i++ ) {
        int n;
        Read >> n;
        nums.push_back(n);
    }

    // now make a Mat from the vector:
    Mat mat1d(nums);
    Mat mat2d = mat1d.reshape(1, 10);

    // your data is in the [0..9] range, so scale up to [0..255] uchar range
    mat2d *= (255/10); 

    cout<<mat2d<<endl; //for testing 
    //
    Mat image; //create an empty image. (you can leave it empty ;)
    //// Apply the colormap, but on the 2d mat, not on the 2d vector, please:
    applyColorMap( mat2d, image, COLORMAP_HOT );
                // Show the result:
    imshow("colormap", image);
    waitKey(0);

enter image description here