Ask Your Question
0

How to detect people with android phone camera(java only)

asked 2013-05-23 12:30:21 -0500

Robert Kilar gravatar image

updated 2013-05-25 11:09:32 -0500

I am looking for algorithms or tools implemented in openCV which would enable to me detecting moving(or not) people on the video(provided in real time by camera of a android phone camera) and count them.

In the other subject I have found:

Quote: Have a look at the LatentSvmDetector. OpenCV already provides different classificators trained to detect people, cars, cats, airplanes..

Maybe some of you have some expierience with detecting people on the video and could provide me some samples. Even if not maybe not you samples but something in the web I did not find.

edit retag flag offensive close merge delete

Comments

were you able to detect a human body on android phone? I tried to change classifier but it didn't work. I got following error at line InputStream is = getResources().openRawResource(R.raw.haarcascade_fullbody)

Error : haarcascade_fullbody cannot be resolved or is not a field

Ankit gravatar imageAnkit ( 2013-05-30 07:49:15 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
1

answered 2013-05-23 15:23:21 -0500

Guanta gravatar image

updated 2013-05-23 15:35:33 -0500

Maybe this documentation helps you further: http://docs.opencv.org/doc/tutorials/objdetect/cascade_classifier/cascade_classifier.html#cascade-classifier , it uses a trained cascade classifier for the frontalface (haarcascade_frontalface_alt.xml) and for the eyes (haarcascade_eye_tree_eyeglasses.xml), replace those two with the trained classifier for the fullbody (haarcascade_fullbody.xml) (note: it is located under <your-opencv-dir>/data/haarcascades/) and you'll detect people in your images.

EDIT Args, just saw that you actually would like to have it for Android, unfortunately I don't have much clue about that, I am not sure if you can just use the C++ functionality. So maybe this post is not useful for you. I'll leave the post for reference.

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

Stats

Asked: 2013-05-23 12:30:21 -0500

Seen: 2,405 times

Last updated: May 25 '13