Object recognition and tracking for Video surveillance applications

Abstract

Object tracking is an emerging concept in the field of video processing. Because of the enormous video surveillance applications, object tracking has received the considerable interest in the recent years. The standard video datasets are collected by the surveillance cameras and given to the researcher for subsequent processing. Because of huge dimension, it provides abundant information than standard images. The thesis focuses on development of a novel algorithm by combining object recognition with feature extraction for tracking in video scenes. The major issue in video sequences is its high dimension. Hence dimensionality reduction is an essential task in all video surveillance applications. Dimensionality reduction can be achieved by feature extraction from the region of interest. Feature extraction methods transform high dimensional video data into lower dimensional space where the originality of the data is preserved. On the other hand, feature extraction selects the most informative regions without disturbing the originality of the data. newline