I have noticed that in the point cloud data of the dataset, the absolute values of the coordinates are very large. In some point clouds, the average absolute value can reach up to 5. However, in the Transforms.json file for the cameras, the distance between the camera and the object center seems to be consistently set to 1.2. This does not seem to be consistent with the original rendering of the point clouds. If I want to match the coordinates of the point cloud with the actual rendering by the camera, does this involve some scaling operations?
I would like to inquire about the scaling operations applied to the coordinates of the point cloud data and how it relates to the camera's rendering process in order to accurately align the two sets of data. Any insights or clarifications on this matter would be greatly appreciated. Thank you!
I have noticed that in the point cloud data of the dataset, the absolute values of the coordinates are very large. In some point clouds, the average absolute value can reach up to 5. However, in the Transforms.json file for the cameras, the distance between the camera and the object center seems to be consistently set to 1.2. This does not seem to be consistent with the original rendering of the point clouds. If I want to match the coordinates of the point cloud with the actual rendering by the camera, does this involve some scaling operations?
I would like to inquire about the scaling operations applied to the coordinates of the point cloud data and how it relates to the camera's rendering process in order to accurately align the two sets of data. Any insights or clarifications on this matter would be greatly appreciated. Thank you!