Ask Your Question

Detect certain object that is partly cluttered

asked 2015-09-16 09:45:39 -0500

215 gravatar image

updated 2015-09-16 16:28:52 -0500

I was wondering if it is possible using haar cascade or some other form of classifier, to detect for instance a skull, when it is partly cluttered with something else, such that when the skull has been detected i would be able to remove only the skull part from the image, and leave out the cluttered part.

An example could be like this

Image found on google

So the shape of the skull is clear, but what i am interested in here is the item stuck into the skull.. Would it be possible vision wise to separate the thing that it stuck in the skull from the skull, or partly do it?

edit retag flag offensive close merge delete



Yes it would be possible. I have experienced that as long as there is like 85% of the object visible you want to detect, that it will still work. So this should work here too. Seperating that thing from the skull is possible I think, but it might need a whole range of options. I think intensity of these scans relate to material, so that could be a starting point.

StevenPuttemans gravatar imageStevenPuttemans ( 2015-09-17 06:19:20 -0500 )edit

How would you remove the skull from the image?

215 gravatar image215 ( 2015-09-17 07:58:31 -0500 )edit

1 answer

Sort by ยป oldest newest most voted

answered 2015-11-24 02:15:42 -0500

belgraviton gravatar image

I think that your task is Anomaly Detection problem. You can read details in paper: Varun Chandola et. al. "Anomaly detection: A survey" ACM Computing Surveys (CSUR), Volume 41, Issue 3, July 2009, Article No. 15

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

1 follower


Asked: 2015-09-16 09:45:39 -0500

Seen: 512 times

Last updated: Nov 24 '15