Ask Your Question

Troup's profile - activity

2016-08-02 10:31:56 -0600 received badge  Nice Question (source)
2016-03-09 03:37:50 -0600 received badge  Popular Question (source)
2014-09-03 06:48:06 -0600 commented question Proper way to unload a loaded classifier?

Hmm. My last question was about -nonsym being removed. Now I want the deprecated cvReleaseHaarClassifierCascade(&cascade). Maybe in a few years I'll catch up.

2014-09-03 04:16:55 -0600 commented question Proper way to unload a loaded classifier?

Thanks for the quick replies - glad I didn't reply too soon before the edit. Yeah, I'd like to make it forget the old one, so I edited my question to better reflect that.

2014-09-03 04:12:53 -0600 received badge  Editor (source)
2014-09-03 03:50:35 -0600 commented question Proper way to unload a loaded classifier?

For instance, if I'm running a classifier on live video and I want to stop the classifier from performing detectMultiScale by taking the xml document I called "load" on, and somehow unload it. I could create a conditional inside processImage to ignore detectMultiScale for whenever I choose, but I'm not comfortable with that solution. If I were to load another classifier, I'm not sure what would happen to the old one, and so I don't want to set myself up for uncertainty.

Thank you for your consideration, berak.

2014-09-02 20:10:36 -0600 asked a question Proper way to unload a loaded classifier?

How do I unload a loaded classifier? I thought by name there may be something like "unload" since I call "load" to begin with, and I do see CascadeClassifier::empty();, but that just checks whether something was loaded (?).

Thanks.

2014-08-26 21:24:11 -0600 commented question TrainCascade: What happened to -sym and -nonsym?

Thank you, Steven.

Matter closed. (uh, when I get enough reputation, I guess)

2014-08-26 03:40:51 -0600 received badge  Student (source)
2014-08-25 16:51:16 -0600 asked a question TrainCascade: What happened to -sym and -nonsym?

Hello:

I see -sym and -nonsym arguments in tutorials for the haartraining utility, but I don't see it mentioned in the new traincascade. I checked the documentation and didn't see it deprecated.

What should we know? Does traincascade now assume symmetry in positive samples, as it was the default in the old utility? Or does it assume non-symmetry by working in a different manner?

Thanks