Object Tracking by Normalized Cross Correlation and PCA Based Template Updating: Comparative Analysis
M. H. Sidram, Nagappa U. Bhajantri · International Journal of Computer Science and Informatics · 2014
The principle behind to detect and track non-stationary object via a sequence of frames is addressed. The proposed strategy pushed the Normalized Cross-Correlation (NCCR) to track object by matching the template and updating the template is encouraged through Principal Component Analysis (PCA). This work remarked with exhaustive experiment and witnessed with comparative analysis over dataset related to outdoor environment. The system kernel reveals the capability to track the object and outcome is fairly acceptable to great extent under different light conditions.