00:02
Here, for the answer, define the px3 matrix that map points in the world coordinate into 2d points in the film place coordination for the left camera p1.
00:18
We can use some perspective project matrix.
00:20
The general form of the perspective project matrix is given by p equals k dot root is multiplied r plus minus r multiplied p closed bracket.
00:38
K is the camera in strict matrix then r is the rotation matrix of the camera, p is the translation vector of the camera.
00:52
Given that the focal length is 1, the camera in intrinsic matrix k for the left camera is a 3x3 identity matrix.
01:02
So the rotation matrix r is determined by the orientation of the camera axis.
01:08
The translation vector p here for the left camera p1 is p, then 1 and 3.
01:19
We saw the perspective project matrix p for the left camera p1 is p, p for p1 is 1 0 0 0 0 1 0 0 0 0 1 0 multiplied dot root is 1 0 0 minus 10 0 1 0 minus 1 0 0 1 minus 3.
01:52
So simplifying this, we would get p p1 1 0 0 0 1 0 0 0 1 minus 7 minus 1 minus 3.
02:14
Now for the essential matrix p here is given that the left camera p1 is the reference camera and p r is the right camera.
02:28
P1 p r the essential matrix is given by e equals p x numeric values i mean algebraic x.
02:45
Here p x is the q symmetric matrix of the translation vector p.
03:11
So the translation vector p from the left camera to the right camera would be p c 1 2 p r which is position of camera p r minus position of camera p 1.
03:36
P c 1 p c r would be equal to 7 minus 10 1 minus 1 2 minus 3...