Hi, I am projecting depth maps to world coordinate using the camera parameters generated by Replicator to get the point cloud from different views. The intrinsic parameters are computed using the following code:
aperture = data["cameraAperture"]
focal_length = data["cameraFocalLength"]
resolution = data["renderProductResolution"]
horiz_aperture = aperture[0]
height = resolution[1]
width = resolution[0]
vert_aperture = height / width * horiz_aperture
fx = height * focal_length / vert_aperture
fy = width * focal_length / horiz_aperture
cx = height * 0.5
cy = width * 0.5
The camera_params.json file contains two matrices: “cameraProjection” and “cameraViewTransform”, which seem to be related to the extrinsic parameters. May I know how to compute the extrinsic parameters using these two matrices?