General Methodology for Implementation of Image Normalization Procedure Using its Wavelet Transform

Image processing is one of the fundamentals of various intelligence systems based on data analysis. In addition among the variety of different methods and approaches to image processing, image normalization methodology can be pointed out. The essence of such image processing methodology is in compensating various geometrical distortions of input image that was obtained through registering test image and forwarding it to the data mining system, in comparison with some reference (master) image. In such case, there arises a problem of time reduction for normalization procedure for some image processing systems. It is suggested that this task can be solved through the use of levels of wavelet decomposition of input image and conducting normalization procedure on separate levels. To solve this problem, this articles studies special aspects of wavelet decomposition of the image into separate levels of decomposition, functionality types of basic groups of image normalization by taking into account the use of wavelet decomposition levels of the test image, general methodology for determining transformation groups for image normalization based on wavelet decomposition. The fairness of the considered methodology is supported by the number of experiments.