strange question about multiple essential matrices

[jonnoj]

In a given video sequence, if the essential matrix Eab between frames a and b as well as Ebc between frames b and c are known, does this information allow for a quicker method of finding Eac  than the direct approach (e.g the 8 point algorithm between frames a and c)?  Presumably the rotational information can be taken from these despite the translational information being partially lost to the two scale invariances.

This scheme is applied across many pairs of frames (i.e finding Eae given Eab, Ebc, Ecd and Ede) so a solution dealing with triplets of frames such as the trifocal tensor is not what I'm looking for unless it also can be easily extended from frame to frame.

apologies if this is the wrong place to ask this.

[Daniel_Thomas_NIUK]

Dear Jonnoj,

If you have any specific questions regarding National Instruments hardware and software then this would be the right place to look. I am afraid in this case that I doubt we will be able to answer your question and I would recommend searching elsewhere.

Good luck.

Kind Regards,

[Hummer1]

You may find what you are looking for in the Calibrtion vision tools.  It provides basis Homographic correction between two sample images .... We have done this using test patterns, but if the locations of fixed points were available in each image, the Calibration should be able to work.  You would need to provide the locations of the known points in image a and the same set of points in image b.  The calibration output is then available for correcting other images...

Hope that helps.