This research project implements a real-time object detection and pose estimation method as described in the paper, Tekin et al. "Real-Time Seamless Single Shot 6D Object Pose Prediction", CVPR 2018. (https://arxiv.org/abs/1711.08848).
MIT License
723
stars
214
forks
source link
Question: The object model must be zero mean? #147
thanks a lot for the contribution to this repository.
I have a question here. To calculate the pose, PnP is used.
R_gt, t_gt = pnp(np.array(np.transpose(np.concatenate((np.zeros((3, 1)), corners3D[:3, :]), axis=1)), dtype='float32'), corners2D_gt, np.array(internal_calibration, dtype='float32'))
The (0,0,0) is concatenated to the eight corner points. So (0,0,0) corresponds to the object centroid? If this holds, then the object mesh must have zero mean?
Hi,
thanks a lot for the contribution to this repository.
I have a question here. To calculate the pose, PnP is used.
R_gt, t_gt = pnp(np.array(np.transpose(np.concatenate((np.zeros((3, 1)), corners3D[:3, :]), axis=1)), dtype='float32'), corners2D_gt, np.array(internal_calibration, dtype='float32'))The (0,0,0) is concatenated to the eight corner points. So (0,0,0) corresponds to the object centroid? If this holds, then the object mesh must have zero mean?
Thanks!