Please use this identifier to cite or link to this item: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4342
Title: CCTV Intruder Detection System
Authors: Chaminda, G.H.A.
Issue Date: 2-Aug-2021
Abstract: With the increasing complexity associated with modern life, people’s concern for safety has assumed top priority in their daily list of needs. In order to ensure their safety and to lead a peaceful life, they have developed and applied a variety of security systems, which have different features and equipment. Normally, there are four major security systems used in important places. They include closed circuit television (CCTV), alarm system, access control system and fire alarm system. Thus, choosing the appropriate system depends on the desired purpose. Among these alternatives, use of CCTV is a popular option adopted by many people in order to ensure the safety of their home front and this has currently been developed from Analog system to IP system. However, at present, the system still lacks the ability to alert users timely and this drawback results in property damages as well as risks to human life. Therefore, in this project, it is proposed to develop an indexing system to find relevant events in a CCTV recording. The whole project consists of four sub models such as Image acquisition model, Image processing model, Information extraction model and E-mail generation stage. The proposed project on CCTV Intruder Detection system can successfully identify the intruders, minimize the false alarms and minimize the human supervision of such systems. It automatically detects the unauthorized activities on the premises: for example, it detects unauthorized access to the premises and sends the warning message to the system owner. Then the system owner can respond according to the incident by remotely logging on to the system and viewing CCTV live stream, playing an alarm, calling the police, etc. This system is highly recommended for the home surveillance which has become an integral part of home security system. The technologies used in this project are OpenCV 3.4.1 and Python 3.0. The project has a distinctive edge in terms of time and efficiency when compared with similar systems discussed here. That is, in an ordinary surveillance system, in case of an incident, all the recorded videos should be manually viewed in order to find the intrusion. This consumes lot of time. In contrast, in the present system, an immediate detection can be obtained because of the email that is generated for the user giving such information as time and image of intrusion. Further, the system has another advantage as it is developed to minimize false alarms. This is achieved through adjusting the threshold pixel value of the image of the intruder. The user can also watch the intrusion through a live video as a follow up and this is another outcome of the project.
URI: http://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4342
Appears in Collections:2019

Files in This Item:
File Description SizeFormat 
2016MIT009.pdf918.57 kBAdobe PDFView/Open


Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.