Abstract
A median model and an improved median model were proposed for background image generation and vehicle detection. The median model has an impressive performance in handling slow moving or even stationary vehicles. A mask-classified updating method was introduced to update the background image in the short-term, where only classified background pixels are being used for updating. The combination of the improved median model and mask-classified updating has several advantages such as it handles slow moving or even stationary vehicles and it can be used in real time image processing.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 168-176 |
| Number of pages | 9 |
| Journal | Proceedings of SPIE - The International Society for Optical Engineering |
| Volume | 5298 |
| DOIs | |
| State | Published - 2004 |
| Event | Imaging Processing: Algorithms and Systems III - San Jose, CA, United States Duration: Jan 19 2004 → Jan 20 2004 |
Keywords
- Background subtraction
- Object classification
- Vehicle detection
- Vehicle tracking
ASJC Scopus subject areas
- Electronic, Optical and Magnetic Materials
- Condensed Matter Physics
- Computer Science Applications
- Applied Mathematics
- Electrical and Electronic Engineering