Camera calibration add radial distortion

6 views (last 30 days)
NASIR FARMANI
NASIR FARMANI on 20 Dec 2020
Edited: Bruno Luong on 23 Dec 2020
Hi
  1 Comment
Stephen23
Stephen23 on 23 Dec 2020
Edited: Stephen23 on 23 Dec 2020
Original question by NASIR FARMANI on 20th December 2020 and many of their (since deleted) comments (not shown here) retrieved from Google Cache:
Camera calibration add radial distortion.
Hi, i have 20 checkerboard images, and i have to add radial distortion without using MATLAB camera calibration libraries.
I have already estimated the projection matrix P for all images, and i have add radial distortion.
I understood i have first estimated P, after i have to estimate k1,k2 from this equations:
where u,v are the idea projections, uhat, vhat are the actual projection and rd.^2 is:
i actual find k1,k2 from this equation = inv(A' * A)* A' * b
after i have to iterate the all process with the new u,v using newton method for find the new u,v.
someone can help me to write this iterative process for find the ideal undistorted image?

Sign in to comment.

Answers (1)

Image Analyst
Image Analyst on 20 Dec 2020
Do you have the Computer Vision Toolbox? There is an undistortImage() function and undistortpoints() function in it.
  3 Comments
Image Analyst
Image Analyst on 20 Dec 2020
What are the alphas? Angles? And the K's?
Actually I don't think that is distortion. Distortion is where the distance the point moved from the ideal location is a function of radius (distance from optical axis) cubed. So
Rdistorted = Rideal + (someFactor * RIdeal ^3)
Which is essentially what your equations show. Actually they're adding in the higher order term - the 5th order polynomial. There are several optical aberrations that have explicit formulas that can be found in the links below:
The most common monochromatic aberrations are:
There is also longitudinal chromatic aberration, lateral chromatic aberration, etc.

Sign in to comment.

Categories

Find more on MATLAB Support Package for USB Webcams in Help Center and File Exchange

Products


Release

R2020b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!