facebookresearch
preparing point clouds as input
Hello
Thank you for the great package. I am trying to use zed 2 camera to get pointcloud and input it to the model. But the point cloud that I am getting from it is not as clean as the one in demo files (input_pc_sunrgbd.ply). I know that the Z direction needs to be upward, but I don't know what are the good preprocessing steps that are needed to get a good input for the model. I am just getting the pointcloud and randomly chooseing 20000 points and save it in a .ply file. I appreciate it if you could help me with this issue. I have attached a depth image and my own point cloud as well.
Would you mind telling me how to convert kinect2' s raw data (RGB + depth images) to votenet' s input format lilk input_pc_sunrgbd.ply?
