Code understanding in vpnp_calib

[BayRanger]

Hi there,

In the file lidar_camera_multi_calib.cpp， I get stuck to understand the following code in the vpnp_calib, which is for ceres optimization. residuals[0] = ud - T(pd.u); residuals[1] = vd - T(pd.v); Eigen::Matrix<T, 2, 2> I = Eigen::Matrix<float, 2, 2>::Identity().cast<T>(); Eigen::Matrix<T, 2, 1> n = pd.direction.cast<T>(); Eigen::Matrix<T, 1, 2> nt = pd.direction.transpose().cast<T>(); Eigen::Matrix<T, 2, 2> V = n * nt; V = I - V; V = I - V; Eigen::Matrix<T, 2, 2> R = Eigen::Matrix<float, 2, 2>::Zero().cast<T>(); R.coeffRef(0, 0) = residuals[0]; R.coeffRef(1, 1) = residuals[1]; R = V * R * V.transpose(); residuals[0] = R.coeffRef(0, 0); residuals[1] = R.coeffRef(1, 1); could you please explain the motivation to do this calculation,? especially about R and V.f

Bests, BR

[autumn999999]

Hi！Did you understand the meaning of this code? I also have the same question. I’d be grateful if you could help me.

[tdcsu]

I am confused with the same problem, Di you understand the meaning of this code? if true, please give me some suggestions. Best TD

[tdcsu]

Hi！Did you understand the meaning of this code? I also have the same question. I’d be grateful if you could help me.

please see issue #29