Sorry, I’m still new to programming
A while ago, I noticed a paper use IPPE to calibrate lidar and camera with kitti dataset
SOIC: Semantic Online Initialization and Calibration for LiDAR and Camera
2003.04260.pdf
They say they have 3D points and corresponding 2D points
And in Figure 4. Show all 3D points and the plane created with RANSAC
How do they use IPPE?
Is to use all 3D points (inlier points) on the RANSAC plane and the corresponding 2D points, then use cv2.solvePnPRansac(np_3d_inlier, np_2d_inlier, camera_matrix, dist_coeffs, cv2.SOLVEPNP_IPPE)?
I use point3D on RANSAC plane and corresponding 2D points in solvePnPRANSAC(flags:IPPE) ,but the reurn rvec and tvec is wrong.
Or can you give me some suggestions how they used IPPE to get extrinsic parameters of LiDAR and camera?
I will be very grateful,Thanks!