Object detection demo program

Program uses AdaBoost Runtime to detect objects in video.

Requirements
- Installed OpenCV 1.0 and LibXML2 libraries
- Camera (or video file)
- Waldboost classifier

Usage
demo -c classifier [options]

Options:
-c filename       Classifier filename (waldboost format)
-t x              Set last stage threshold to x (default 0.0)
-s x              Set classifier window shift to x (2)
-f x              Set pyramid scale factor to x (1.33)
-l x              Set number of pyramid levels to x (8)
--video filename  Use video instead of camera input
--output filename Save output video to file (XviD codes is osed when available)
--size w h        Set video frame size to w x h px (512 384)
--linear          Use linear interpolation (off)
    
demo -c test.xml -t 4.5 --size 320 240 --linear --output test.avi

If video file is not specified camera is used instead.

When program starts classifier is loaded and user can see detections in video marked by white rectangle. Performance informations are displayed on stdout - frames per second and average features evaluated per sample.
