Adaptive Video Processing for Object and Event Recognition
Adaptive Video Processing for Enhanced Object And Event Recognition in UAV (Unmanned Air Vehicle) Imagery :
  • Apply adaptive object and event processing techniques to increase detection probability and decrease false alarm rate;
  • Recognize and describe events in UAV video defined by a change in the relationship between objects over time.
frame regions
Building Recognition by Consistent Line Clusters (demo)
Lines are classified into color, orientation, and spatially consistent clusters and consistent line clusters are applied in the application of building recognition.
 
Building Recognition by line intersections (demo)

Boat Recognition (demo)


Vehicle Recognition (Demo)
The objective of this project is to recognize the vehicles in a given image.  The picture below is an example of the results I have gotten.  The whole procedure includes three steps: 1. vehicle horizontal position localization; 2. Vehicle width determination; 3. Vehicle vertical position localization.  Three methods were implemented: 
  • Symmetry-based recognition
  • Intensity-based-symmetry recognition
  • Horizontal-edge-based recognition
See my demos (java applet) of all these methods. 
  • Yi Li, Vehicle Recognition, Project Report, 1999.