Rotation and Scale Invariant Intrusion Detection for Security Systems

Rupali N. Wadekar, Dr. M. S. Nagmode

Abstract: There are some areas where a very tight security is needed. These areas include Line of Control, Airports, etc. Also, in some areas, it is not possible to appoint human soldiers due to the worst environmental conditions and high risk of attack by enemies. In such cases, a complete independent system is necessary which will automatically detect the intruding objects and destroy them. Thus, such a system will ensure tight security without endangering the precious life of human soldiers. The system proposed here is a complete independent system in which camera, image processing software, supplementary hardware takes care of destroying intruding objects automatically. The system is kept simple from implementation point of view. Total 20 objects of different sizes and shape are tested. From experimental results it is found that system gives good performance for various shapes and sizes. Thus the system is scale invariant as well as rotation invariant. Overall recognition accuracy is 93 %.

Keywords: Feature extraction, MATLAB, object recognition, rotation invariant, scale invariant

How to Cite?: Rupali N. Wadekar, Dr. M. S. Nagmode, "Rotation and Scale Invariant Intrusion Detection for Security Systems", Volume 4 Issue 6, June 2015, International Journal of Science and Research (IJSR), Pages: 295-300, https://www.ijsr.net/getabstract.php?paperid=1061501, DOI: https://dx.doi.org/10.21275/1061501