Problem with stitching photoes into a panoramic image

[Jian QI]

HI

There are four fisheye photos,i want to stitching them with hugin(version 2018.0.0,windows),but it turns out to be very bad.I tried stitching them with PTGui Pro on windows,it works fine.so i want to figure out how can i handle these photos with hugin well like PTGui?

(the lens of photos is circle fisheye and fov is 200,i am not very sure now..)

I have tried for a long long time, and sorry for my poor english

[Jesse Kennedy]

Here's what I came up with. I had to manually place the control points. The assistant couldn't do it automatically. You can improve the quality by adding and tweaking control points and optimizing variables. 

--
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 hugin-ptx+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/aace0a04-14e2-454f-acd2-5b66bca2d278%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[panostar]

On Tuesday, March 20, 2018 at 7:46:17 AM UTC, Jian QI wrote:

HI

There are four fisheye photos,i want to stitching them with hugin(version 2018.0.0,windows),but it turns out to be very bad.I tried stitching them with PTGui Pro on windows,it works fine.so i want to figure out how can i handle these photos with hugin well like PTGui?

I got a decent result using Hugin 2010, but using 2018 was unsuccessful as it failed in the enblend process.

John

[T. Modes]

Am Dienstag, 20. März 2018 08:46:17 UTC+1 schrieb Jian QI:

(the lens of photos is circle fisheye and fov is 200,i am not very sure now..)

In this case you need probably a bigger fov of 380. Hugin/panotools is measuring the fov about the full image width, and your picture circle does only cover a small area. Therefore the big value is needed. With the bigger value cpfind is also finding control points.
There is another issue with this image set: the optical center of each image is at different positions (switch between the images and you will see how the picture circle jumps). So you need to unlink the image center shift parameters d and e and optimize them. (Another way would be to assign each image an own lens. Then is this also given.)