I want to calibrate a line scan camera which captures the image of a flat moving object. I want to avoid the successive application of a non-linear model to describe the lense distortion and a linear model to describe the projective transformation.

Therefore, I fitted a polynomial of order 3 or 4 to describe the deformation of the image points in x and y direction, respectively. It seems to work.

Does anyone know a better model?

The Brown-Conrady model does not describe the deformation very well. Interestingly, the lense distortion smoothly changes during the line scans.

Thank you in advance

More Matthias Demant's questions See All
Similar questions and discussions