Transformation of camera coordinates (study notes)

Assuming a point Pw in the world coordinates,
the external parameters of the camera-the pose is known (R, T), then the point P(X, Y, Z) = (RPw+t) in the camera coordinate system can be obtained, and then normalized After processing, get the coordinates P1 (X/Z, Y/Z, 1) under the normalized plane, and then multiply P1 by the camera's internal parameter K to get the coordinates (u, v) of the pixel plane
K=
[fx 0 cx
0 fy cy
0 0 1]