Related Experiment Videos
Recovery of image blocks using the method of alternating projections
Jiho Park1, Dong-Chul Park, Robert J Marks
1Baylor University, Waco, TX 76798-7356, USA.
Summary
Recovery of Image Blocks using the Method of Alternating Projections (RIBMAP) effectively restores missing image blocks in block-based coding. This novel technique achieves superior spatial structure recovery and higher peak signal-to-noise ratios compared to existing methods.
Area of Science:
- Digital Image Processing
- Signal Processing
- Computer Vision
Background:
- Block-based image and video coding are prone to block-loss artifacts.
- Existing restoration methods often struggle with faithful spatial structure recovery.
Purpose of the Study:
- To develop a novel technique for block-loss restoration in image and video coding.
- To improve the accuracy and fidelity of restored image blocks.
Main Methods:
- Developed Recovery of Image Blocks using the Method of Alternating Projections (RIBMAP).
- Employs orthogonal projections onto constraint sets in Hilbert space.
- Utilizes extracted vectors from surrounding blocks to identify matching spatial information.
- Incorporates three convex constraints for continuity and imposes adherence via alternating projections.
Main Results:
- RIBMAP demonstrated faithful recovery of spatial structure in simulations.
- Achieved a peak signal-to-noise ratio (PSNR) of 34.65 dB on the "Lena" JPEG image with 25% block loss.
- Outperformed existing methods including hybrid edge-based, projection-based, and DCT-based interpolation algorithms.
Conclusions:
- RIBMAP is a highly effective algorithm for block-loss restoration in digital images.
- The method offers significant improvements in image quality and structural integrity over prior techniques.