How to perform image registration on non-rectangular image

Question

Jeremy Eekhoff 2022-5-31

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1730935-how-to-perform-image-registration-on-non-rectangular-image

编辑： Walter Roberson 2024-11-8，6:15

I'm interesting in peforming image registration using imregtform on a non-rectangular region of interest that would be the "moving" image. Is there any way to do this? I initially tried to turn all the pixels within the "moving" image matrix that I don't want to affect registration to NaNs, but the imregtform function won't accept that input.

3 个评论
显示 1更早的评论隐藏 1更早的评论

Jeremy Eekhoff 2022-6-6

That would not work because the black pixels would still affect registration. My images do not have a black background.

raym 2024-11-8，6:09

编辑：Walter Roberson 2024-11-8，6:15

Dear Jeremy D. Eekhoff,

I'm sorry to trouble you on on the forum here.

I read your article "Three-dimensional computation of fibre orientation, diameter and branching in segmented image stacks of fibrous networks" by Jeremy D. Eekhoff and Spencer P. Lake.This is a really great application on detection of fibers in 3D space. I want to have a try and download the code from github: https://github.com/jeremyeekhoff/Fiber-Analysis-Algorithm .

I used a demo 3D stack fiber stainning image to run the function fiberAnalysisAlgorithm.m on R1026a, with MinGW64 compiler on win8.1.

However, it shows the error on the two mexw64 files:

Invalid MEX-file 'mexfilename': The specified module could not be found.

I have tried using dependencywalker to find which dlls are missing. It first says that several dlls in C:\Program Files\MATLAB\R2016a\bin\win64 is missing. Thus I add this folder to env and restart computer and re-run, then it says the following dlls are missing, although most of them exist in folders below, except API-MS-WIN-CORE-SHUTDOWN-L1-1-1.DLL.

API-MS-WIN-CORE-KERNEL32-PRIVATE-L1-1-1.DLL

API-MS-WIN-CORE-PRIVATEPROFILE-L1-1-1.DLL

API-MS-WIN-CORE-SHUTDOWN-L1-1-1.DLL

API-MS-WIN-SERVICE-PRIVATE-L1-1-1.DLL

I searched and find that API-MS-WIN-CORE-SHUTDOWN-L1-1-1.DLL is obselete in Win8.1

https://www.dll-files.com/api-ms-win-core-shutdown-l1-1-1.dll.html

After I download API-MS-WIN-CORE-SHUTDOWN-L1-1-1.DLL and place in the same folder of m code, the error still persist, and

dependencywalker shows more complicated errors. I guess it is a problem of mexw64 compiling environment differs too much with my environment. Could you give me some instructions on this?

Thank you.

----the path to most of dlls above:

C:\Windows\System32\downlevel\API-MS-Win-Core-Kernel32-Private-L1-1-1.dll

C:\Windows\SysWOW64\downlevel\API-MS-Win-Core-Kernel32-Private-L1-1-1.dll

C:\Windows\WinSxS\amd64_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_2c5177c069a919dd\API-MS-Win-Core-Kernel32-Private-L1-1-1.dll

C:\Windows\WinSxS\x86_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_d032dc3cb14ba8a7\API-MS-Win-Core-Kernel32-Private-L1-1-1.dll

C:\Users\rmd\AppData\Local\GeoGebra_6\app-6.0.583\API-MS-Win-Core-Kernel32-Private-L1-1-1.dll

-----

C:\Windows\System32\downlevel\api-ms-win-core-privateprofile-l1-1-1.dll

C:\Windows\SysWOW64\downlevel\api-ms-win-core-privateprofile-l1-1-1.dll

C:\Windows\WinSxS\amd64_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_2c5177c069a919dd\api-ms-win-core-privateprofile-l1-1-1.dll

C:\Windows\WinSxS\x86_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_d032dc3cb14ba8a7\api-ms-win-core-privateprofile-l1-1-1.dll

C:\Users\rmd\AppData\Local\GeoGebra_6\app-6.0.583\api-ms-win-core-privateprofile-l1-1-1.dll

-----this one may be missing, I will copy it and have a try

C:\Users\rmd\AppData\Local\GeoGebra_6\app-6.0.583\api-ms-win-core-shutdown-l1-1-1.dll

-----

C:\Users\rmd\AppData\Local\GeoGebra_6\app-6.0.583\api-ms-win-service-private-l1-1-1.dll

C:\Windows\System32\downlevel\api-ms-win-service-private-l1-1-1.dll

C:\Windows\SysWOW64\downlevel\api-ms-win-service-private-l1-1-1.dll

C:\Windows\WinSxS\amd64_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_b8233abb5511544f\api-ms-win-service-private-l1-1-1.dll

C:\Windows\WinSxS\x86_microsoft-windows-m..namespace-downlevel_31bf3856ad364e35_6.3.9600.16384_none_5c049f379cb3e319\api-ms-win-service-private-l1-1-1.dll

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Matt J 2022-6-6

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1730935-how-to-perform-image-registration-on-non-rectangular-image#answer_979925

编辑：Matt J 2022-6-6

If you have the Optimization Toolbox, you could use lsqcurvefit (don't be misled by the name) to minimize your own reigstration cost function, one which accepts an ROI mask.

5 个评论
显示 3更早的评论隐藏 3更早的评论

Matt J 2022-8-1

编辑：Matt J 2022-8-1

What is usually done to improve robustness is that you first smooth and downsample the images and perform an initial registration. Then you reduce the smoothing and repeat the registration with the previous registration result as your initial guess. This is repeated until the registration is done at full resolution.

Jeremy Eekhoff 2022-8-1

Ok thanks, I'll try that out.

请先登录，再进行评论。

Answer 2

Image Analyst 2022-8-1

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1730935-how-to-perform-image-registration-on-non-rectangular-image#answer_1019120

normxcorr2_demo.m

Is it rotating and scaling while it translates? If not, you could use normalized cross correlation. See attached demo.

Or try one of the panoramic stitching algorithms. They find matching regions, which is necessary to construct the overall image.

Panorama and Image Registration - MATLAB & Simulink

Feature Based Panoramic Image Stitching - MATLAB & Simulink

Panoramic photograph stitching

Cris’ Image Analysis Blog | Panoramic photograph stitching

5 个评论
显示 3更早的评论隐藏 3更早的评论

Jeremy Eekhoff 2022-8-11

Hey - quick following up on this. I was able to somewhat get this working using fmincon, but the results weren't quite as good as I get using imregtform and I would often have to change the optimizer options for different sets of images. I had the idea to try to go back and use imregtform and set the pixel intensities of all pixels outside of the polygonal region of interest to uniformly distributed random values, the idea being that the blocks of random pixels would have zero mutual information with the fixed image regardless of where it ended up and therefore wouldn't affect the registration. This ended up working better than my attemps to make my own registration optimizer. Does this seem like a reasonable solution or am I missing something that could be problematic?

Matt J 2022-8-11

Does this seem like a reasonable solution or am I missing something that could be problematic?

I can't argue with success, but I vaguely wonder why setting the background pixels to some constant wouldn't do the same thing. An atomic distribution is independent of every distribution as well.

请先登录，再进行评论。

How to perform image registration on non-rectangular image

3 个评论
显示 1更早的评论隐藏 1更早的评论

回答（2 个）

5 个评论
显示 3更早的评论隐藏 3更早的评论

5 个评论
显示 3更早的评论隐藏 3更早的评论

另请参阅

类别

标签

产品

版本

Community Treasure Hunt

How to perform image registration on non-rectangular image

3 个评论 显示 1更早的评论隐藏 1更早的评论

回答（2 个）

5 个评论 显示 3更早的评论隐藏 3更早的评论

5 个评论 显示 3更早的评论隐藏 3更早的评论

另请参阅

类别

标签

产品

版本

Community Treasure Hunt

3 个评论
显示 1更早的评论隐藏 1更早的评论

5 个评论
显示 3更早的评论隐藏 3更早的评论

5 个评论
显示 3更早的评论隐藏 3更早的评论