Abstract: Today security in the public places is the main concern issue faced by world. Visual surveillance is used to analyze object behavior. This paper presents an approach for abandoned object detection in security surveillance videos using raspberry pi, a single board computer and python is used as programming environment. It involves static object, moving object detection and video tracking to identify the events occurring in the scenes. The proposed system takes video input, processes video with adaptive background modelling, morphological operation is performed to find the abandoned object and motion of owner of object. A warning is initiated on the screen when abandoned object is detected. Experiments are performed on PETS 2006 and ABODA dataset.

Keywords: Surveillance, Raspberry Pi, Python, Adaptive Background Modelling, Abandoned Object

| DOI: 10.17148/IJIREEICE.2018.674