Moving object detection in a sequence of images taken from non-stationary camera
Moving object detection is a vital aspect of motion analysis. It has drawn an increasing attention in the recent years due to its applications such as in communication, traffic monitoring, security surveillance, robot navigation and servoing. Despite the fact that much research efforts have be...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2004
|
Subjects: | |
Online Access: | http://eprints.uthm.edu.my/7707/1/24p%20NORAN%20AZIZAN%20CHOLAN.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Moving object detection is a vital aspect of motion analysis. It has drawn an
increasing attention in the recent years due to its applications such as in communication,
traffic monitoring, security surveillance, robot navigation and servoing. Despite the fact
that much research efforts have been devoted to this area, detecting moving object using
non-stationary moving camera remains a great challenge. The research undertaken in this
thesis is mainly concentrated on developing a reliable and robust detection system which
incorporates some operation on images such as thresholding, blob labelling, blob
matching, filtering and blob analysis. The basic idea behind this system is that the motion
of the moving object is di fferent with the motion of background object. Path transversed
within a certain period of observation of the moving object is usually longer than
background object. By using blob labelling and blob matching operation, this system
would be able to track binary blobs over an arbitarily long image sequence. The criteria
for matching binary blobs from two adjacent frames are position, height, width, area,
colour and aspect ratio. If the the path transversed ofa binary blob within a certain period
of observation is sufficiently long, then the tracked blob is considered as moving object. |
---|