Hi All,
I am trying to generate the 3D corneal surface from the Zernike Polynomials. I am using the following steps, can anyone please let me know whether they are accurate
Step 1: Converted the cartesian data (x, y, z) to polar data (rho, theta, z)
Step 2: Nomalised the rho values, so that they will be less than one
Step 3: Based on the order, calculated the Zernike polynomials (Zpoly), (for example: if the order is 6, the number of polynomials is 28 )
Step 4: Zfit = C1 * Z1 + C2 * Z2 + C3 * Z3 + ......... + C28 * Z28
Step 5: Using regression analysis, calculated the coefficient (C) values
Step 6: Calculated the error between the predicted value (Zfit) and the actual elevation value (Z)
Step 7: Finally, converted the polar data (rho, theta, Zfit) to Cartesian coordinates to get the approximated corneal surface
Thanks & Regards,
Nithin
Seems fine to me. Step 6 will tell you if you did it right. The residual should be small, and also shouldn’t show any low order structure that seems too similar to any of the fit zernikes.
- Badges
- Science topic
- Similar topics
- Psychology