point cloud texture mapping

I am currently doing a 3d scene reconstruction from stereo images in Matlab. The camera calibration, disparity map estimation and point cloud generation work fine. My problem is that I can only correctly map the left image onto my point cloud. I want to find a way to map the right image. The reason why I'm not just using the left image as texture is that I have two different cameras, RGB and IGB, and I would like to map the IGB data onto the point cloud. As far as I understand. The point cloud in Matlab is stored as an (M x N x 3) array, where (M,N) describes the pixels in the (left) image. I understand that I cannot colour each point separately based on its xyz coordinates, but instead use the indices M and N and the colour information stored in the image. But since there is a rotation and translation between the two cameras, the the same (M,N) pair does not describe the same xyz point in both images. So I have to somehow project the right image onto the left image. I've done some research, and narrowed down the problem to a Homography transformation, which should do exactly what I want, right? I used the computeRectificationParameters function (found inside StereoParametersImpl.m), to calculate the transform. It seems to project the images onto a common plane, but I can't figure out how to calculate the correct translation so that the left and right images overlap.

So my question is: How do I correctly incorporate the translation into the Homography matrix?