Share Email Print

Proceedings Paper

Morphological scene change detection for night time security
Author(s): Benjamin Jarvis; Andrew J. Tickle
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Morphological Scene Change Detection (MSCD) systems can be used to secure environments by sensing potential intruders and alerting security personnel to any security risks. To achieve this, the system compares the input from a camera to a reference image quantifying the level of change between the images, raising the alarm if this change is greater than a set triggering level. Morphological operators are than used to reduce the effect of any image change not related to a potential security risk; this includes noise and other minor changes thus decreasing the risk of false alarms. However in low light conditions MSCD systems can fail due to the reduced intensity differences between images containing security threats and reference images. This paper documents a proof of concept for a system that would use night vision images to address this problem. Here a low light scope camera attachment is used in place of a night vision camera and shows modifications to the previous MSCD system, which improves the performance when used with night vision images. The analysis of the modified system’s performance in different low light environments, this includes analysis of appropriate binary threshold and alarm triggering levels for a range of environments. The latter includes indoors at a distance, indoors at close range, outdoors at a distance and outdoors at close range. The results shown demonstrate that MSCD systems operating in low light conditions have the potential to be used as a useful tool in a security system and are compared to the original to demonstrate the improvement.

Paper Details

Date Published: 19 October 2012
PDF: 11 pages
Proc. SPIE 8540, Unmanned/Unattended Sensors and Sensor Networks IX, 85400J (19 October 2012); doi: 10.1117/12.974758
Show Author Affiliations
Benjamin Jarvis, Coventry Univ. (United Kingdom)
Andrew J. Tickle, Coventry Univ. (United Kingdom)

Published in SPIE Proceedings Vol. 8540:
Unmanned/Unattended Sensors and Sensor Networks IX
Edward M. Carapezza; Henry J. White, Editor(s)

© SPIE. Terms of Use
Back to Top