An enhanced noise reduction model for template matching using asymmetric correlation technique

Template matching is widely used in image processing. In symmetric normalization both the compared vectors are normalized in same way, and in filter design, only the template is modified and then is correlated with the reference image. The above mention is a problem from, to overcome this we moved to asymmetric correlation. Noise reduction in image using asymmetric correlation by template matching, provides an efficient and noise strong template matching method based on Asymmetric correlation (ASC), where symmetric correlation provides some issues towards template matching. DFT calculation has been reduced and calculated for only a selected pixel, which provides ASC to be more efficient and time as been saved. We show that this asymmetric normalization is more robust to noise than other cross correlation variations, such as the correlation coefficient. This follows a nonlinear normalized image with a linear normalized image to have an accurate set of image as an output.