Author : B.Dhananjaya, B. Rama murthy and P.Thimmaiah

Video tracking in real time is one of the most important topics in the field of surveillance systems. Detection and tracking of moving objects in the video scenes is the first step in the information extraction in many computer vision applications. In this paper , an intelligent method for object detection and tracking in real time video using OpenCV on ARM-9 Beagle Bone Black is explored. Processing a video stream to segment foreground objects from the background is a critical step in many computer vision applications. Background subtraction is a commonly used method for achieving this segmentation. Gaussian Mixture-based Background Segmentation Algorithm and morphological operations are used in this paper for object detection and tracking. In the first step a video is taken as input , is divided in to frames , each frame is converted to binary frame then applying background subtraction algorithm to detect the moving objects. Erosion and Dilation Morphological operations are performed on these frames to remove the unwanted shadow, tracking and labeling the moving objects. The main processing unit is a ARM Cortex A8 processor based Beagle Bone Black (BBB) with Linux (Ubuntu) operating system installed with OpenCV. The indexing of identified objects and automatic labeling enables the developed system suitable for surveillance applications.

Keywords: Background subtraction, OpenCV, ARM – Cortex, Morphological Operations , Object tracking , surveillance.

Article published in International Journal of Current Engineering and Technology, Vol.5, No.2 (April-2015)