skippy130's profile - activity

Hi,

I'm using OpenCV 3.2 in Android Studio to help build an app that does facial detection and recognition. The issue I'm having right now is that the facial detection is really wonky. What I mean is that often times in the video feed it'll identify the ceiling as a face and make a jumpy rectangle that keeps going in and out and sometimes just disappears even when the only thing in the frame is a face.

So far I've only used the frontal_face.xml file to do facial detection but I've been told that I can use additional haarcascade xml files, like frontal_face_alt.xml, together to yield better detection. However I have no idea how to use two or more cascades at once. If anyone has any advice on how to use at least the frontal_face and frontal_face_alt cascades together please let me know. Or if you have any other ideas that could improve facial detection then please feel free to share that too.

Thanks in advance!

I'm using version 2.4.11.

I'm basically trying to make a database, stored in the app, that will be used to recognize individuals. I also want to have the option for an individual that isn't recognized to be added to the database and thus be recognized in the future.

Also, do you know of any sort of documentation for the openCV Java wrapper? Or is it pretty much a one-to-one conversion of syntax from C++ to Java?

Hi,

I've recently got OpenCV up and running in Android Studio. I'm trying to make a basic app that does some facial recognition but when I try to use createEigenFaceRecogizer or FisherFaceRecognizer I get a syntax error basically stating that Android Studio has no idea what that function is.

Looking online I've seen people mention that the Java wrapper excluded the facial recognition functions for some reason and that a work around is to create your own wrapper.

So I guess my first question is if this is true? Second, if that is true, then could anyone guide me on how to go about making a wrapper to make those functions work? I have no idea on how to go about this at all, so even a link to a tutorial about the very basics of making a proper wrapper would be greatly appreciated. Sorry if this is a too vague or uninteresting question, but I appreciate any help that you can offer.

Thanks! Looks like setThreshold() doesn't exist in Python but I found setDouble('threshold', value) using the help(cv2.createEigenFaceRecognizer()) function. I'd give you an upvote but it looks like I can't yet, but thanks a lot for the help!

Hi,

I'm trying to set the threshold of a facerecognizer object that I've made in Python. Looking online I found that you can use the getDouble('threshold') method to get the threshold and the set('threshold', doubleThresholdValue) to set it.

The first method worked just fine but when I try to use the second one I get an error 'AttributeError: 'cv2.FaceRecognizer' object has no attribute 'set''.

So I was hoping that someone could point me in the right direction in terms of methods available to set the threshold.

And is there any documentation for openCV that exists specifically for Python or other languages like Java? I've been using this (I apparently can't post links, sorry) as my resource but it seems like it's geared towards C++.

That documentation is fine and I haven't had any real trouble translating it to Python but I'm wondering if there might be more problems like this in my future if these methods don't have a one-to-one correspondence between different languages. Thanks in advance!