I want to use find-2d-object to do object detection and do 3d pose estimation with the point cloud coming from 3d lidar. i impelented object detection, but donot know how to do 3d pose estimation with the point cloud of lidar.
how can i do that? would please give me some points?
Re: 3d pose estimation with the pointcloud of lidar
To re-use the simple 3D pose estimation form find_object_2d node, you need correspnding camera_info and depth image. To generate the depth image from the lidar, you need proper TF tree and the node pointcloud_to_depthimage could be used to generate the depth image.