opencv - People Detection and Tracking -


i want pedestrian detection , tracking.

input: video stream cctv camera.

output:

  1. #(no of) people going left right
  2. # people going right left
  3. # no. of people in middle

what have done far: pedestrian detection using hog , svm. detection decent high false positive rate. , slow running in android platform.

question: after detection how calculate required values listed above. can tell me tracking algorithm have use , algorithm pedestrian detection.

or should use tracking algorithm? there way without it?

any references codes/blogs/technical papers appreciated.

platform: c++ & opencv / android.

--thanks

this somehow close research problem.

you may want have this website gathers lot of references. in particular, work done group oxford present therein pretty close doing, since using hog detection. (that work has been extremely illuminating me). epfl , julich have work done in field.

you may want give this review describes several detection/tracking techniques, involving variants of hog algorithm.


Comments

Popular posts from this blog

user interface - Python attempting to create a simple gui, getting "AttributeError: 'MainMenu' object has no attribute 'intro_screen'" -

jquery - Common JavaScript snippet to share files on Google Drive, Dropbox, Box.net or SkyDrive -

Android Gson.fromJson error -