3

The information matrix is just the inverse of the covariance matrix. I recommend you read the page I linked, or just google covariance matrix. Essentially it contains how certain you are in your measurements. (The lower the number the less uncertain you are). As an example: The translation matrix between nodes(ignoring rotation for now). Your covariance ...


1

The simplest explanation will be: In structure from motion, it estimates structure(xyz points), camera locations, camera intrinsic. In graph optimization, it only estimates camera locations. In the graph SLAM, the structure is just a by-product of a corrected trajectory or graph nodes. E.g. implementing Bundle adjustment with g2o -> You can do it by ...


1

The challenges you are going to face is more how to mount the cameras and find the relative positions between them (also the rotation). You can find a lot of designs for mounts for the cameras, also 3D printers' designs. The calibration is pretty similar to the one for one camera chess pattern etc. There are also a lot of tutorials explaining how you should ...


1

Yes you can use two camera the same as using single stereo camera for depth perception. Step on calibration camera for two camera and stereo camera is same.


1

If the cameras are stationary it should work to use Structure from Motion (https://github.com/mapillary/OpenSfM). Failing that the cameras are stationary you could attempt to create a factor graph in gtsam to solve for the camera movement, object movement, and projections between the cameras and object. See a simple SFM setup here: https://gtsam-jlblanco-...


Only top voted, non community-wiki answers of a minimum length are eligible