Fred schrieb:
You won't find any ready-to-use code because RANSAC is quite simple in
its form but requires to set arbitrary thresholds which are very
application dependant. If all you want to do is fit a straight line when
there are some outliers robust regression (use google) methods (like
minimizing the median error instead of the mean error or applying a
running median filter on the data first) are easier to implement because
less parameter tuning is required.