Lab 4 Detection of humans in a video stream - meetsriharsha/RTBDA_5543 GitHub Wiki

#Welcome to the RTBDA_5543 wiki!

09/21/2016

In this lab exercise, we developed an application that is capable of tracking humans in the given video stream.

In the main program (ProcessingPanel.java) we created a Detector object to store the identified human frames. We used the fullbody template from the already available BuiltInCascade class of HarrCascadeDetector. Later we generated the rectangles for the human region. These rectangles were added to each frame and displayed in the output video in green color.

The input video consists of people in different poses and the application will scan through each frame of the video and highlights the identified humans with a green color bordered rectangle. Sample test output panopto video link is below:

Human Detection - Panopto Video

##Sample Output Screenshots: