An Edge-Preserving Signal-Correlation-Based (EP-SCB) Post-Processing Scheme for CFA Image
| Autor: | Yang, Yi-Hong, 楊翊弘 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 106 In this dissertation, we propose an edge-preserving(EP)-modified signal-correlation-based (SCB) post-processing, called EP-SCB, as an enhancement to any existing interpolation method. The proposed Bayer image reconstruction system thus consists of two operational phases. The first phase performs an initial estimation of the missing RGB colors by using an existing interpolation method, while the second phase applies EP-SCB post-processing. Since a certain class of images may not fulfill the premise of having small variations in local color difference, which is assumed by SCB-type interpolation, thereby resulting in a deterioration in composite peak signal-to-noise ratio (CPSNR) after EP-SCB post-processing, a threshold test on local variance ratio is also devised to conditionally switch off the second phase. Experimental results show that the EP-SCB post-processing with variance ratio test gives a worse average CPSNR than the original interpolation methods in none of the Kodak and IMAX image groups experimented. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|