We have developed an algorithm to detect salient motion in complex environments. The effectiveness of  algorithm to robustly detect salient motion is demonstrated for a variety of real environments with unimportant motions due to lighting changes, swaying branches, flowing or rippling water, and camera shaking or vibration. The details can be found at paper [1]. All demo videos are in MPEG1 format.
 

Top left:  Original video, 
Bottom left: motion, 
Top right: temporal difference, 
Bottom right: detected salient motion.

1. Swaying leaves (video 408KB)

2. Lighting changes (video 2MB)

3. Rippling water (video 3.6MB)

4. Fountain and waterfall (video 1.6MB)

5. Camera moving (video 543KB)

 

[1] Robust Salient Motion Detection with Complex Background for Real-time Video Surveillance

Ying-li Tian and Arun Hampapur
IEEE Computer Society Workshop on Motion and Video Computing, Breckenridge, Colorado, January 5 and 6, 2005 PDF

Other Research Areas:

• Robust Background Subtraction
• Object Classification 
• 2D Tracking 
• 3D Multi-Person Tracking 
• Articulated Human Body Tracking 
• Active Head Tracking
• Coarse Head Pose Estimation 
• Position Independent Absolute Head Pose Estimation
• Face Cataloger
• Video Privacy
• Multi-scale Tracking & Index Browser
  
• Real Time Alerts
• Middleware for Large Scale Surveillance (MILS)
• Performance Evaluation of Surveillance Systems