Paper Title
Human Tracking using Template Matching Methodology
Abstract
Human tracking system has become a very important where security is the priority. Tracking the object in the
frames from the video is a key research topic in the computer vision community. The proposed system consist of targeting
human face of interest, where we investigate long-term tracking of that face in a video clips. We are using Template
Matching Algorithm which is novel tracking framework (TLD) that explicitly decomposes the long-term tracking task into
tracking, learning and detection. The tracker follows the object from frame to` frame. The detector localizes all appearances
that have been observed so far and corrects the tracker if necessary. A P-N learning method estimates the errors by P-expert
which estimates missed detections, and N-expert which estimates false alarms. The learning estimates detector’s errors and
updates it to avoid these errors in the future. Experimental results shows that our approach achieves good performance on
video sequences.
Keywords - Template, Learning, Detection, Bounding Box, Convolution, diffeomorphism.