Ultrahigh Accuracy Camera Calibration

Opticist.org has developed an ultrahigh-accuracy camera calibration algorithm.  The performance is considerably better than the widely used Matlab camera calibation toolbox and OpenCV camera calibration function. The function has been incorporated into our program: http://opticist.org/node/73

The instruction can be found here: http://faculty.cua.edu/wangz/publications/advanced_camera_calibration.pdf

FAQ can be found here: http://www.opticist.org/node/338

Sample images can be downloaded on this page after logging in.

If you use our program, please cite our paper so that more people can learn and use this technique:

M. Vo, Z. Wang, L. Luu, and J. Ma, “Advanced geometric camera calibration for machine vision,” Optical Engineering, Vol. 50, No. 11, 110503, 2011.

This project is supported by US Army Research Office under grant W911NF-10-1-0502.

Projecting Points

Hi Emrah,

As I said earlier, my code is very rudimentary and it is just to show that the software actually works. Here are a few things you need to consider:

(1): Lens distortion. I ignored them in my code for simplicity. Yet, the distortion definitely has non-negligible effect!

(2): Difference in starting index of Matlab and C++. Matlab starts from 1 instead of 0. So, the displaying code needs to be modified.


Projecting Points

Hi Minh,

Thank you for your answer.

(1) I undistorted the image by Moire Software before drawing the points on it. Isn't that suppose to give the exact result?

(2) I have tested the "+1" coordinates. Actually, since the error on image is not fixed but growing in x direction, that didn't help, neither.

Thank you again for discussing. 



Dear Admin,

thank you for your replay. I tested your software with your default setting and now it works.

I have another question. Do I have set  the autofocus's camera on when I take the pictures?

because the calibration gives me different results.

Thanks again for your help.


I have some problem with identification of the  control points.

Could I have the defoult settings for the check-board control points,

I am usining your pattern (  7 control points for each rows and 5 for each colon, the distance between two ccontrol points is 40.6 mm)

Are these values correct? Why the identification is made just for one or two pictures in 25?

Thank you very much for your assistance.

Calibration pattern


If you used the patterns we attached together the software, there will be 12 control points in the row and 9 in the column. The distance between them is 20.32mm.



The manual looks very helpful. Wonderful!

is camera calibration program available?

Dear Admin 

when in the fall 2011 will be the camera calibration program with ultrahigh accuracy will be available?

Please help me .

thank you .




Dear Admin 

whenever I am trying to calibrate the camera with the software it shuts down!!

I have the ring circled images taken by a SONY HD camera .

I have taken 30 images.

I tried with 30 and again with 4 images also . 

But after a while it shuts down and doesnt give me the calibration parameters.

I appreciate your help.

Thank you.



We have never experenced such thing before. One possible reason is that you are running it on old model computer that cannot restrain the computational cost of our program.

BTW: it will be best if you can upload your images so that we can test them.

2DIC or 3DIC

Dear Admin

I appreciate for your help.

I ahve a quick question.

Does it take too long for the 2D/3D DIC analysis of an image?

Thanks .



Sharif: It depends on your images. If there are clear correspondence between them (good speckle pattern), the processing time will be as fast as 3000 points/sec depending on the criteria and other parameters used.

Dear Admin

Dear Admin

Thank you very much  for your quick reply.

But I have tried the software with my new HP(pavilion dv6) intel core i7.

While running the program it hangs up all on a sudden and shuts down the program.

Yes , I do have the images. They are in .bmp extensions. I tried with the circular images that you have attached with the software in the zipfile ( took the pictures with a SONY HD handy camera). 

Each of the bmp image is almost 9MB. Can you please let me know how to attach the pictures in this post. 

I couldnt find any attachment option in it like as yahoo or hotmail . If you could please let me know how to attach the .bmp pictures .

Thank you again very much .

I appreciate your help.


Upload images.

Sharif, can you upload it to some servers and upload that the link to this website?

Btw: try to run the software again under administrator mode.

Dear Admin

Dear Admin

Thank you for your prompt reply.

I tried to run the program being an administrator but dint work.

This is what I am getting as a message:"Moire MFC application has stopped working(heading of the box) , A problem caused the program to stop working correctly. Windows will close the program & notify you if a solution is available."

I have uploaded 4 out 30 files (as they are size is larger).

Thank you again very much.

I appreciate your help.









1. Can you double check your images? I can download only the 1st one. 

2. You need to convert your images to grey scale format. This can be done by going to Tools-> Convert True Images to 256...

3. The result would be better if you can make flatter pattern.

Dear Admin

Dear Admin 

Thank you for your prompt reply and time.

I worked according to your advice and this time the program did not hang. I tried with 4 of the images and converted the true images to 256...

This program ran and gave the message at the end :"This file is shown here for you to check the reprojection error.

The 1st col. is the board # index (valid boards only), and the 2nd col. is the reproj_errs (smallest to largest)


02 7.112419e-002

03 7.582088e-002

01 8.642542e-002

04 9.256300e-002


Mean: 8.148337e-002 "

I think I did not get the calibration parameter yet and the program is mentioning about the projection error.

Any suggestions?

I appreciate your help.

Thanks again.


Release note


Please look at our release note for further interpration. We did show how to get the parameters in it.

Thank you Dear Admin

Dear Admin

Thank you very much.

It worked and I went through the release note and found details about it.

I got the results in two rows and each row consisted 30 parameters .

Is it like one row signifies if the origin of image pixel coordinate lower left corner and the 2nd row if world coordinate?

I appreciate your help.

Thank you very much.


Release note


Those 30 values are the results of 1 set of calibration. The fact that you saw 2 rows meant that you calibrated twice. The 1st row shows the most recent calibraion result. The 2nd row shows the calibration parameters that you calibrated before.

Please note that the first 2 parameters, which represent the focal length in pixels of x and y direction respectively, are negative. This is due the our choice of coordinate system as described in the release note.

Dear Admin

Dear Admin

I have a quick question .

Is there anyway we can use these camera parameters and retrieve the 3D depth of an image ?

Or the cartesian co-ordinates(x,y,z) for a point in an image?

I appreciate your help.

Thank you .


Thank you

Thank you again.

And thank you very much for your prompt reply.


Camera calibration

Hi Sharif,

We will make our latest version online on Dec 1st, 2011.


Is the camera calibration software available?

Its 2nd december,2011.

Is the camera calibration software available on www.opticist.org?


Camera calibration

Hi Sharif,

We are sorry for the delay as I was too busy to write up the description for the program.

The new version, Moire 0.950, is now available. Please check the admin's post for details.


Thank you .

Thank you .


Thank you very much.


I want to know that in which day of this summer Opticist.org will release the camera calibration program and when you will release the 3D-DIC,thank you!!!

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer