对图像用小波进行层小波分解（Wavelet decomposition of image using wavelet）.doc

对图像用小波进行层小波分解（Wavelet decomposition of image using wavelet） Introduction: This paper starts with the two-dimensional wavelet theory and analyzes and processes its application in image processing, in an attempt to show that wavelet analysis has its unique features in image processing. This article describes the following points: First, the basic concept of wavelet. Image compression, image denoising, image enhancement, image smoothing processing Two: wavelet basic concepts Definition: set the wavelet transform for Fu Liye, when meet the condition that the perfect reconstruction condition or identical resolution conditions. When we called a wavelet or wavelet, the generating function through extension and translation, too. We call it a wavelet sequence. Among them, a is a stretching factor and B is a translation factor. Wavelet transform is a signal of the time - scale analysis method, it has the characteristics of multi-resolution analysis, and can be used to describe the local characteristic of signal in time-frequency domain two, is a fixed window size but its shape variable, time-frequency localization time window and the frequency window variable analysis method. The frequency and time resolution is the low frequency part is high, the high frequency part has higher time resolution and low frequency resolution, it is suitable for detecting the transient abnormal phenomena with the normal signal and its component, so called signal analysis microscope. Wave analysis is to decompose the signal into two parts: low frequency Al and high frequency dl. In the decomposition, the lost information in low frequency Al is captured by high frequency dl. In the decomposition of the next layer, Al is decomposed into two parts: low-frequency A2 and high frequency D2. The lost information in low frequency A2 is captured by high frequency D2, and so on, it can be decomposed further. The two-dimensional wavelet function through one-dimensional wavelet transform of the tensor product