Real Time Object Detection & Tracking over a Mobile Platform

Background/Objectives: Nowadays, object detection and tracking is an important issue in robotics and computer vision systems, especially in video surveillance, robot navigation and autonomous vehicle navigation. Methods: In this paper, we propose a fast method for object tracking and recognition within the context of a mobile robot acquiring real time images from a top mounted IP camera. The aim of the proposed method is to let the robot identify multi-targets within a scene, then move toward the desired object. The method is based on a new vectored contour to identify objects from HSV images. Findings: Experimental results have shown that our method best fitted the mobile platform and gave excellent competitive results in real time tracking. Our proposed method has shown better adaptation com-pared to SURF and other state of the art tracking methods, especially in terms of time and simplicity, specifically when the camera is not in front of the target object, i.e. at different inclination angles and distances. Application: The object detection and tracking proposed in this work can be implemented on many fields such as video surveillance robotic navigation or in industry in classification of objects depending on their forms or colors.