Real time human detection in video streams

Detecting humans in films and videos is a challenging problem owing to the motion of the subjects, the camera and the background and to variations in pose, appearance, clothing, illumination and background clutter. We develop a detector for standing and moving people in videos, testing several different motions coding schemes and showing empirically that orientated histograms give the best overall performance. The resulting detector is tested on several databases including a challenging test set taken from feature films and containing wide ranges of pose, motion and background variations, including moving cameras and backgrounds. Experimental results from real time video are provided. These results are validated on a challenging test set containing 1000 human examples. Key-Words: Human detection, video streams, HOG, SVM, Haar fatures , ROC