TRACKING PERSON IN OUTDOOR AND DYNAMIC ENVIRONMENT
No Thumbnail Available
Date
2012-11-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This research is intended to track person in the video sequences. The videos used in this research
consist of persons walking in the outside of building with dynamic environment. There are two
cameras used in this research with different field of view but with overlapping views. Each video
has specifications i.e. frame size of 576x768 pixels, rate of 30 frame per second, and avi format.
After reading the input, we do background estimator. The background is taken from the initial
frame where there is no people walking in the observed area. After the segmentation process, then
we use Kalman Filter to predict the tracked person walking in the view. The filtered position
yielded by the above algorithm is used to mark the rectangles which state the foreground. When
there is occlusion in one view, the homography from this view to other views is estimated from
previous tracking results and used to infer the correct transformation for the occluded view. The
experiment results show that the tracking process in outdoor and dynamic environment can be done
correctly, although some problems still occur to be a challenge for the next research.
Description
Keywords
Outdoor environment, person tracking