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
I would like to run this software in a scenario where some fixed or less fixed object is showing up in the camera and LIDAR data, which should be ignored for calibration purposes, e.g. camera an LIDAR are mounted on large machinery which moves them around, or stuff is hanging into the image/lidar FOV.
Where would I add logic to filter out pixels or points so they do not participate in the optimization?
The text was updated successfully, but these errors were encountered:
I think such a mask can be implemented by modifying the NIDCost class that measures the normalized information distance between a point cloud and an image.
BTW, I suppose just filling masked regions with a uniform color (e.g., black or white) should work in most cases because the NID effectively ignores such outliers (at least in theory).
I would like to run this software in a scenario where some fixed or less fixed object is showing up in the camera and LIDAR data, which should be ignored for calibration purposes, e.g. camera an LIDAR are mounted on large machinery which moves them around, or stuff is hanging into the image/lidar FOV.
Where would I add logic to filter out pixels or points so they do not participate in the optimization?
The text was updated successfully, but these errors were encountered: