WAVELET SHRINKAGE: AN APPLICATION TO DENOISING
| Autor: | Patrick J. Van Fleet |
|---|---|
| Rok vydání: | 2019 |
| Předmět: | |
| Zdroj: | Discrete Wavelet Transformations ISBN: 9781119555414 Discrete Wavelet Transformations: An Elementary Approach with Applications |
| DOI: | 10.1002/9781119555414.ch6 |
| Popis: | This chapter presents a basic overview of wavelet shrinkage and its application to signal denoising. It discusses two wavelet‐based methods used to denoise signals: the VisuShrink method and the SureShrink method. VisuShrink utilizes the wavelet shrinkage algorithm with a universal threshold. The chapter looks at two examples that implement VisuShrink. The first example illustrates the method on the test signal. This example is somewhat artificial since people set the noise level when they form the noisy vector. The second example introduces the idea of image segmentation. In this application, pixels are separated into different regions based on similarity. Denoising the image before segmenetation is applied can sometimes increase the effectiveness of the algorithm. The SureShrink method utilizes Stein's unbiased risk estimator. It calls for a tolerance to be computed for each highpass portion of the wavelet transformation. A test is performed on each highpass portion to determine its sparseness. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|