Correlation of robust Haar-like feature

Presented is a new template matching method using correlation based on robust Haar-like features. The main result of this reported work is the robust computation of Haar-like features to limit the influence of outliers and any effect that may cause a deviation from a perfect match. Simulation results show that the correlation of robust Haar-like features outperforms the correlation of Haar-like features, phase correlation, and the normalised cross-correlation while having comparable results with orientation correlation and robust correlation of the second derivatives.