Analysis of Wavelet Packet Tree Decomposition to Create Image Pattern to be Used in Image Classification System

Digital data and the most widespread due to their use in many vital and important applications for humans, especially in classification and image recognition systems. Due to the large size of the color image, it is pereferable to represent the color image by a unique pattern consisting of a small number of elements. In this paper research we will introduce a wavelet packet tree decomposition as a good method og extracting images patterns. We will study and analyze verious approaches of using wavelet packet tree decomposition, some recommendation will be done in order to fix the pattern length and make it capapale to be used correctly in a classification tool.