On Sun, Sep 18, 2022 at 12:50 PM 黄禄轩 <[email protected]> wrote:
> in theory, TrX, TrY and TrZ can solve the problem, but in my case sometime > it goes wrong and i don't know why. > Optimizing TrX together with Yaw, and starting from one of those very far from correct, tends to go wrong. Photos taken with near zero yaw still have some yaw because you can't aim the camera perfectly (unless it is mounted on a track). So you probably need to end up optimizing both TrX and Yaw. But you should not try to optimize them together until they are each close to correct. For a camera mosaic, 0.0 is close to correct for Yaw and not for TrX. So optimize TrX with Yaw set to 0.0, then optimize them together starting from the result of that first optimization. The same is generally true of pitch vs. TrY, so those should be done the same way. But in the described linear sequence of pictures, there is less room for the pitch vs. TrY ambiguity to do much harm (might as well do pitch/TrY correctly since it is important to do Yaw vs. TrX correctly). Roll and TrZ don't have similar ambiguity vs each other, but TrZ may need to be done at the same time as TrX for good results. With a well distributed collection of control points, I think including Roll with TrX, TrY, TrZ initially might do better than adding it later. But either way ought to work pretty well. The one really important reason to optimize twice is to get TrX approximately correct before allowing nonzero Yaw. -- A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ --- You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/CALe0Q_%3D-8rD8gWq93k%3DizmCvhkrZDNbqMHWuU2AYhViQrsVVnA%40mail.gmail.com.
