1
$\begingroup$

I have an assignment on a time series that contains the x positions of a point in space. This series comes with another one of the same length, where every element goes from 0.0 to 1.0 and represents the accuracy of the measurement with the same index: 1 is perfect accuracy, while 0 is missing/completely garbage data.

I'm already doing linear interpolation on the points where the accuracy it's exactly 0, but afterwards I'm just using a median filter and a Gaussian filter, ignoring the accuracy values. I wonder, is there some kind of filter that can take into account the accuracy data and "smooth more" the series when the accuracy is lower (and the noise is higher)?

Improve this question
$\endgroup$
1
  • $\begingroup$ Maybe useful context: the accuracy is never actually 1 $\endgroup$
    – ThePirate42
    Commented Dec 10, 2023 at 23:43

1 Answer 1

Reset to default
1
$\begingroup$

Since you have accuracy data, this seems like it could be a good application for an alpha-beta filter, or some variation of that. Since you have accuracy data, you would be able to adjust the filter gains according to the accuracy of that sample.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.