Performance Enhancement through Handling of False Classification for Smart Video Surveillance
Lavanya Sharma1,* and Nirvikar Lohan2
1,*Department of Computer Science, Uttarakhand Technical University, Dehradun, India.
2Department of Computer Science, College of Engineering Roorkee, Roorkee, India.
Over the last decennium, Visual surveillance has become an active research domain for academicians, researchers or industry due to its rapid day by day growing importance in terms of realistic environment. The solutions of the video surveillance are security tools that help us to monitor various things in terms of moving object (i.e. locations, monuments, building, people, etc. In this paper we proposed an efficient method for detection of object using background subtraction technique by enhancing the exiting method. To preserve the shape and removal of noisy pixels some post processing tools were also used. Comparative analysis of our method with considered state-of-the-art method reveals that proposed method shows better outcomes both in terms of qualitative and Quantitative analysis.
Keywords: Video Surveillance, Background Subtraction, Object Detection and Tracking, Pedestrian, Morphology.
