< Previous | Contents | Manuals Home | Boris FX | Next >
Before the solve, you can run the Create Planar Corners script to augment each selected planar tracker (2-D or 3-D) with four automatically-generated clones, one at each corner of the planar tracker. This gets you a lot closer to having enough for a traditional 3-D solve. (Naturally, this assumes that the 3-D planar tracker is solid.)
Because these trackers are all interrelated, they are not as data-rich as five independent trackers, but should be counted as 3-4 for purposes of assessing whether there are enough for a traditional 3-D solve (6+). Two planar trackers should be enough for a basic 3-D solve, as long as the two trackers are not co-planar. In a practical situation you might use several planar trackers and several regular trackers for a 3-D solve.
If a planar tracker is small, or much larger than the pixel region actually being tracked, and its corner data is not accurate, do not run the Create corner trackers script, and use only the nominal tracker center in the solve.
If you have a high degree of confidence in the field of view values produced from your planar 3-D trackers—and they are all the same—you can configure the camera lens to Known mode (which you may have already done to make multiple planar trackers match), so the 3-D solve will use it too. More on that in a moment.
Note : a solver-computed FOV based on many trackers and frames will be more accurate than a FOV you have manually determined from one planar tracker on one frame.
©2023 Boris FX, Inc. — UNOFFICIAL — Converted from original PDF.