You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, I want to get a point cloud from a depth scene image. I use the given camera parameters, but the obtained point cloud has certain distortion. Can you explain it?
The text was updated successfully, but these errors were encountered:
Our depth images are actually measuring Euclidean distance to the camera, not planar depth. See #9 for more details. If you're using depth images to obtain a point cloud, you need to take this nuance into account. And to be 100% correct, you will also need to take into account our per-scene camera intrinsics. See here for more details.
But actually none of the above steps are necessary. We release position images, and you can simply reshape these images from [H, W, 3] to [H*W, 3] and concatenate them together to obtain a consistent world-space point cloud.
Hello, I want to get a point cloud from a depth scene image. I use the given camera parameters, but the obtained point cloud has certain distortion. Can you explain it?
The text was updated successfully, but these errors were encountered: