matlab - 3D Correspondences from fundamental matrix -
In MATLAB, I have calculated using the normalized (two images) from that to the same image point in the 3D space is required. To do what I understand, I need the rotation and translation of the camera's image. The easiest way of the course will be to take the images first, but it will be a huge obstacle for my application because this extra step will be required.
I mention the mention, though it seems that initial translation and rotation are required, so that I think a calibrated camera is necessary or my understanding is decreasing.
So my question is how do I automatically rotate the translation / translation so that I can do the image iterated / tri-in the 3D space. Any MATLAB code or pseudocode will be fantastic.
You can use the original matrix to retrieve the camera matrix 3D digits from your images Triangular However, you must know that you will be rebuilt, a projective will be rebuilt, and Euclidean will not be one. This is useful if your goal is to projective inorientes to measure the original scene like cross resistanction, line intersection, etc., but this angle and The distance will not be enough to measure (you have to check the camera for that)
If you have access, then you can check it at Section 9.5.3 where you will need to go from the original matrix to a pair of camera matrix, which will allow you to calculate a projective rebuilding (I assume that the same content appears in section 6.4 of the book of the mother). Since, you have the function vgg_P_from_F, vgg_X_from_xP_lin, and vgg_X_from_xP_nonlin.
Comments
Post a Comment