I read the term "robust (image) segmentation" a lot in literature but I couldn't find a definition of the term. Based on the literature, I can guess the term may mean that its segmentation result is accurate enough for further analysis with the result, regardless of all different image conditions that are possible in a given application field. I want to hear from experts.


1 Answer 1


Robustness is generally considered to be good (acceptable) performance under uncertainty. Generally, this uncertainty is "noise" of some description but it could be some other form of uncertainty.

This paper by Antoine Vacavant says:

the definition of robustness by Peter Meer as “An estimator is considered robust only when the estimation error is guaranteed to be less than what can be tolerated in the application”.

I doubt you'll find a definition of the term or, rather, I suspect you'll find many definitions of the term. Possibly as many definitions as there are authors.

In robust control, the idea of robust is different:

Probably the most important example of a robust control technique is H-infinity loop-shaping, which was developed by Duncan McFarlane and Keith Glover of Cambridge University; this method minimizes the sensitivity of a system over its frequency spectrum, and this guarantees that the system will not greatly deviate from expected trajectories when disturbances enter the system.


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.