Tell me what refactoring is (3) The results of refactoring

Theoretically, after decomposing the image and then reconstructing it, the reconstructed image should be different no matter which layer it is from. According to your intention, after the processing and decomposition is completed, the reconstructed result will be the same as before. , what else should we deal with? There are two types of wavelet reconstruction. One is complete reconstruction, which is to reconstruct the final decomposed maximum-order approximation information and each order of detail information together to obtain the original signal. Usually, for some purpose in application, Reduce certain levels of information to reconstruct. For example, when denoising, you can remove several low-level detailed information and completely reconstruct it using high-order details and maximum-order approximation information to obtain the original image denoising result; the other is single-level reconstruction, which is Regardless of the approximation or detail wavelet coefficients, they cannot be directly applied, and it is usually necessary to reconstruct the wavelet coefficients to obtain the processing results. Since the wavelet coefficients of each layer of decomposition are different, the reconstructed image is of course different. Your basic concepts are completely confused!