opencv - Structure from motion from multiple views -

I am using OpenCV to extract 3D models from multiple views (pictures). I've got a condition where inputs are two images and the results are 3d digits that rebuild 3D models.

I want to add more than two images to complete reconstruction of all the scenes. Therefore, I have 3d digit given pictures 1-2 and a 3D digit illustration is 2-3.

Any suggestions?

Assuming you are rebuilding 3D using stereo matching algorithms, between images 3D rebuilding 1 results in 3D coordinates of images in 1-2 coordinate systems. Similarly, 3D rebuild between images in 2-3 results in 3D image 2 in the coordinate system points.

So you just have to change the 3D coordinate system of your second point cloud of image 1 for image 2 to one. It uses rotation matrix R and image translucent between 1-2 vector t.

Edit : Note that this method of merging two point clouds is very basic, and you can improve the accuracy of 1-2-3 images at a time Using 3D rebuilt (like bundle adjustment) using I do not think this view is available though in OpenCV.