Object Tracking in Intelligent Video Surveillance System Based on Artificial System
Mukesh Choubisa, Vijay Kumar, Mukesh Kumar, Samrat Khanna
Abstract
Object moving tracking is one of the furthermost attractive and critical chores in computer science and computer vision research domain. Object moving tracking is underpinned for various higher-level automated applications in different research domains such as surveillance of video, motion capture and Augmented Reality (AR) in computer vision. Object tracking is the significant section of an intelligent video surveillance system (IVS) which can be further used for finding many other apprehensive or suspicious activities. Through the independent and exhaustive research of the current object motion and moving tracking technology, collective with hands- on application of IVS (intelligent video surveillance). In computer science many different algorithms available for object path tracking, object tracking and object motion detection, but mean shift algorithm (MSA) is the most popular and basic approach. We use mean shift algorithm (MSA) for our research work. The proposed work in this paper with following objectives, a) Video series frame production for tracking of a moving body, b) Design of experiments for object detection, c) Formulation and implementation of an algorithm to detect object tracking.The algorithm was applied on an image frame of a recorded video. Based on multiple input images, we were able to detect the object and track its successive image frames. The realistic execution may want to obtain the object body tracking without losing any input frame and could magnificently coat bounding field. After completion of the mean shift algorithm (MSA) it would be able to produce an image sequence. The maneuver and precision of the algorithm was tested by successfully overlay function in the identified object and user defined object.