OpenCV equivalent for Matlab's rdivide? - matlab

Question

For example we have expression using rdivide in Matlab:
B = bsxfun(#rdivide, A, A(4,:));
How can we write equavalent expression for opencv?
Opencv has divide function, but seems it can't be used for matrix with different dimensions:
Mat t1= Mat::ones(2,3,CV_64FC1);
Mat t2= Mat::ones(1,3,CV_64FC1);
Mat dst;
divide(t1,t2,dst);
this don't work, so we need to replicate one row to matrix to match dimensions of t1 or use divide with 1 row in cycle.
My solution for opencv(A modified inplace):
for(int i=0;i<A.rows;++i)
{
divide(A.row(i),A.row(3),A.row(i));
}
Is there any simpler way?

Answer 1

You can use the repeat function of OpenCV to replicate a matrix.
The equivalent OpenCV code for the above mentioned MATLAB command is following:
cv::Mat B = A/cv::repeat(A.row(3),4,1);

Answer 2

In addition to #sgarizvi solution, you may find this wrapper to Matlab rdivide helpful:
#include <opencv2\opencv.hpp>
#include <iostream>
using namespace std;
using namespace cv;
Mat rdivide(const Mat& A, const Mat& B)
{
int nx = A.cols / B.cols;
int ny = A.rows / B.rows;
return A / repeat(B, ny, nx);
}
Mat rdivide(const Mat& A, double d)
{
return A / d;
}
int main()
{
Mat1f A = (Mat1f(3, 5) << 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
Mat B = rdivide(A, A.row(2)); // Divide by matrix, works also for cols: e.g. A.col(2)
Mat C = rdivide(A, 2); // Divide by scalar
cout << "A: " << endl << A << endl << endl;
cout << "B: " << endl << B << endl << endl;
cout << "C: " << endl << C << endl << endl;
return 0;
}