215 International Journal for Modern Trends in Science and Technology Abandoned Object Detection Nilesh Ghuge 1 | Pravin Dhulekar 2 1 PG Scholar in E&TC (VES), SITRC, Nashik, Maharashtra, India 2 Assistant Professor, Department of Electronics and Telecommunications, SITRC, Nashik, Maharashtra,India. To Cite this Article Nilesh Ghuge and Pravin Dhulekar, “Abandoned Object Detection”, International Journal for Modern Trends in Science and Technology, Vol. 03, Issue 06, June 2017, pp. 215-218. Now a day, security cameras have been installed at a high rate in places where there are extensive grounds and large crowd is there. Surveillance system has become an important aspect in security and a necessity to keep proper check. There was a time when video surveillance was mostly used by the government and big companies. But now the scenario has changed. Presently the use of surveillance camera has increased andit is increasing more. The number of cameras installed for security has been increasing year by year. The main reason is security enhancement including the prevention of incidences of terrorism. Therefore, we proposed a method which detects abandoned objects and algorithms can be used to assist security officers monitoring live surveillance video by directing their attention to a potential area of interest. KEYWORDS: Object detection, infrequently moving objects, visual attention, visual surveillance, background model Copyright © 2017 International Journal for Modern Trends in Science and Technology All rights reserved. I. INTRODUCTION In video surveillance tasks, many times it is required to isolate foreground objects of interest, from background [1]. Based on the extracted foreground objects, high-level tasks, like identifying target objects [2] and distinguishing activities from videos, can be addressed more effectively. Assuming that the camera is stationary, motion plays a major role in video-based foreground/background separation: foreground objects are usually moving while the background is relatively static. Many approaches, like optical flow and background subtraction, have been developed to track the motions of foreground objects, based on which the foreground and the background can be separated. Optical flow requires that the foreground objects move all the time. However, in practice, foreground objects may show sporadic motions, i.e., objects stay static for a long time and have short-duration motions occasionally, e.g., abandoned objects [3], removed objects [4], and persons stopping for a while and then walking away [5], [6]. Background subtraction is another type of effective approaches that have been widely used to identify the moving foreground objects from a clean background [7], [8], [9], [10],[11],[12]. In this paper, we target to monitor the objects. It is assumed that there are abandoned objects such as a bag or a box. Then there is a possibility where things are intentionally left behind. It can be explosive materials which aim at public facilities. If we can detect them, it is possible to prevent the terrorism. For instance,“Detecting unattended packages through human activity recognition and object association”. This study uses fixed point camera. However, fixed point camera cannot detect far objects. Assuming that the camera is stationary, motion plays a key role in video- based foreground/background separation: foreground objects are usually moving while the background is relatively static. So, to guard from any unpredicted situations from occurring we need the surveillance system to ABSTRACT International Journal for Modern Trends in Science and Technology Volume: 03, Issue No: 06, June 2017 ISSN: 2455-3778 http://www.ijmtst.com