TEMPLATE MATCHING BY WIENER FILTERING

In this paper, a template matching method based on Wiener filtering is introduced. The basic principle is that the template image acts as a blurring function, the matching coefficients signifying whether the given template exists or not in the corresponding position. When the blurring function and additive noise act on matching matrix, the searched image is produced. Then template matching is translated to image restoring. Wiener filter is used to restore the matching coefficients matrix. Compared with related methods, this method exhibits superior performance in localizing the matching points and also shows robustness.