Morphological based moving object detection with background subtraction method

Abstract

Moving object detection is an active research area in the field of video processing and computer vision forming the base of many video analytic applications. The typical impediments in detection of moving objects including dynamic scenes, sudden illumination variations, complex background, effects of shadows, bootstrapping, video noise and camouflage receive the attention of researchers around the globe. This study proposes a morphological based approach for moving object detection. Morphological operations are combined with background subtraction technique and thresholding for experimental purpose. Furthermore, this paper outlines the methods of moving object detection and summarizes the recent research trends in this direction. The goal of this research is to explore the effects of morphological changes on the detection of moving objects. The preliminary results indicate that the proposed approach can generate accurate and complete moving object keeping the required details intact for meaningful object detection.