All geek questions in one place

Opencv - face recognition used to detect objects without a face?

Can (or maybe should) use face recognition software to detect faces without a face? For example, let's say I'm trying to find a can of soda in a picture of a bedroom. If I use haartraining algorithms in many photos of soda cans, will the face recognition function find the capacity of the soda in the bedroom picture?

+4
source share
2 answers

Yes, you can. These algorithms are based on extracting functions from training sets. If you walk past them, it will be good to detect facial features. If you give them the cans, they will be well versed in the cans.

Another idea is that you specialize the algorithm on a specific object, limiting the search, but hari is not.

+6
source

I think you mean face recognition, not face recognition.
In general, face recognition works on a different look or texture in the face area. They also implicitly accept a mostly flat object.

You can use the same classifiers to detect other predominantly flat objects that differ in appearance, for example, facial features, such as eyes and mouth, book covers, etc.

Soda cans are cylindrical, so you may need to use several overlapping rotated views of each can to make them work.

+1
source

Source: https://habr.com/ru/post/1433502/

More articles:


All Articles