Question - best way to get my hands on a solver with 5-parameter warp modeling?

[KyleJosling]

Hi Dima,

Thanks for this really awesome camera calibration toolkit.

I'm working on improving a camera calibration process and we have integrated your tool and made process changes according to your recommendations - we have had really good success with this.

Unfortunately I got a new calibration board and I think its warped asymmetrically (or maybe the print is scaled weird) - bummer.

I'm working on getting it mounted flatter by using a backing, but wanted to explore my options for using the 5-parameter warp modeling for this board. I see the richer-board-shape branch is quite stale - is my best option trying to take these changes and try to put them on to master? or do you have some other branch?

Thanks!

[dkogan]

Hi. I think you should fix your board. The richer-board-shape code does work, and I used it for a period of time for a project, when I had no other option. But it's a bit of a hack. There's zero documentation other than version control, and I haven't thought about that code in a long time. I want to say that there were drawbacks to that implementation, but it has been so long, I don't remember what they are anymore.

If you really need it, you should look at that branch, figure out how it works, and see if you can run it. If you want to be really cool, you can try to merge it into master. I'm doing other things at the moment, but feel free to ask questions.