Image Restoration with Perturbation in Basis Expansion
| Autor: | Chiou, Jyh-Yuion, 邱智勇 |
|---|---|
| Rok vydání: | 1996 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 84 In the area of image processing, there are several algorithms proposed to solve this image restoration problems.In this thesis, a robust approach restoring the blurred image is presented. The scheme is derived based on the perturbation methodwith the concept of basis expaansion to minimize a equilibrium function which is gotten from a regularization process.The relationship of the desired solution variables and original pixel domain image signal is clearly stated in teh discretization and the concept of signal representation bya set of the weighted basis. From the basis expaansion, we can transform the signaaal from the pixel domain intothe coefficient domain and reduce the effect of the error propagation during the updating process.Thus we can achieve the effect to increase the performance of the restored image both in the convergence rate and visual system.to enhance the performance of the restored image which is passing through the perturbation methos alogorithm, an equalizeris also proposed to enhance the contrast of the restored image.Thus, the restored imaage will be clearer than the image without image enhancement. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|