Application of Wavelet Transform in Image Matching Technology
Ya Zhang, Wen Shi · Applied Mechanics and Materials · 2013
Aiming at the image process demands, a kind of image matching recognition algorithm based on statistical characteristics of the image after wavelet transform was proposed. This algorithm adopts wavelet to perform multiple scale two-dimensional wavelet transform for image. The corresponding progeny images energy distribution in different direction and different scale will be taken as statistical characteristics. The targets to be recognized can be identified through the angle's cosine value between the extracted statistical features from the targets and the statistical characteristics of the standard library image. Finally, the bar codes of books are adopted to verify this method and the results show that the algorithm has good recognition effect for the different characters in bar codes.