Robust Statistics
At an early stage, science students learn that averaging is an effective way of eliminating noise and improving accuracy. However, Chapter 3 demonstrated unequivocally that median filtering of images is far better than mean filtering, both in retaining the form of the underlying signal and in suppressing impulse noise. Robust statistics is the subject of systematically eliminating outliers from visual or other data. This appendix seeks to give useful insights into thisimportant subject.
Look out for:
• the concepts breakdown point and relative efficiency.
• M-, R-, and L-estimators.
• the idea of an influence function.
• the least median of squares (LMedS) approach.
• the RANSAC approach.
• the ways these methods can be applied ...
Get Machine Vision, 3rd Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
