Video Surveillance for Sensor Platforms by Mayssaa Najjar Milad Ghantous & Magdy Bayoumi

Author:Mayssaa Najjar, Milad Ghantous & Magdy Bayoumi
Language: eng
Format: epub
Publisher: Springer New York, New York, NY

(5.2)

This approach is rather fast. But it has high memory requirements as it requires buffering n frames to obtain good approximations of the background. Just like the previous technique; it does not accommodate for a statistical description of the scene as needed in the case of swinging trees for instance, and does not update the subtraction threshold [22]. Approximated Median Filter (AMF) estimates the median without keeping a large buffer [23]. The running estimate of the median is incremented by one if the input pixel is larger than the estimate, and decreased by one if smaller. Eventually, the estimate converges to the median value. This technique is good for indoor applications and is used for urban traffic monitoring, but suffers from slow adaptation when there is a large change in background. Any error in the background takes a long time to be corrected.

Instead, the Running Average Filter (RAF) uses exponential weighting and selective updating of background pixels. It requires only one frame to be saved i.e. low memory requirements. The background may be initially chosen as the first frame that does not contain objects. It is then updated at each new frame along with the threshold Th t (x,y) as shown below:

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.