Fits an ellipsoid into a 3D set of points, allows some constraints, like orientation constraint and equal radii constraint. E.g., you can use it to fit a rugby ball, or a sphere. 'help ellipsoid_fit' says it all. Returns both the algebraic description of the ellipsoid (the nine coefficients of the quadratic form) and the geometric description (center, radii, principal axes).
evecs is eigenvectors matrix. Following the principal axis theorem, eigenvectors represent principal axes of ellipsoid. Regarding use with magnetometer calibration, those are needed to perform so called "soft iron" calibration, in order to properly shrink the ellipsoid into sphere. I would recommend taking a look at freescale's AN4246 & AN4248.
This is well done. I especially appreciate the multiple input methods.
The flag=0 method can easily return imaginary results for noisy data. Note that the flag=0 method is also trying to find the primary axes of the ellipsoid. For those getting imaginary results when using flat=0 (the default), if your ellipsoid is expected to already be aligned with x, y, and z, then just use flag=1. This works quite well even when flag=0 falls apart.
I see some folks are using this for magnetometer data, which has a "hard iron" offset. To convert from raw data to points on a unit sphere:
[c, r] = ellipsoid_fit(raw, 1);
% For a single point, k:
unit_sphere(:, k) = (raw(:, k) - c)./r
% For multiple points, vectorized:
unit_sphere = diag(1./r) * (m - c * ones(1, size(m, 2)));
% Just for kicks using bsxfun:
unit_sphere = bsxfun(@times, 1./r, bsxfun(@minus, m, c))
I used your code but the results are not correct with my data. The eigen values are very big. Should you normalise the Covariance matrix? I would also want to ask how did you derive from the 9 parameters of ellipsoid, the covariance matrix, center etc.. which book, notes did u use?
Hi, nice and efficient piece of code.
Nevertheless, I believe that the polynomial expression that you compute in a first step is not guaranteed to be one of an ellipsoid. I think that it could be any quadric (hyperboloid, paraboloid, etc), right?
It is probably no problem if you have much data (corresponding to a real ellipsoid) to fit your quadric on.
But in case of sparse data (for instance, data points only on a couple of planes), you could have bad surprises (and in particular negative eigenvalues).
I have a question: we can find the center, the radii of the ellipsoid and also the 3*3 direction matrix "evecs" using this program. Then how do you determine the three rotation angles using "evecs"?
Thanks for your help!