Related Experiment Videos
Single image super-resolution using self-optimizing mask via fractional-order gradient interpolation and
Qi Yang1, Yanzhu Zhang1, Tiebiao Zhao2
1Shenyang Ligong University, China.
ISA Transactions
|April 9, 2017
Summary
This study introduces a new single image super-resolution method using adaptive fractional-order gradient interpolation. The technique enhances image detail and texture preservation, outperforming existing methods.
Area of Science:
- Computer Vision
- Image Processing
Background:
- Low-resolution images lack sufficient data for clear restoration.
- Existing super-resolution methods struggle with artifact-free reconstruction and texture preservation.
Purpose of the Study:
- To develop a novel single image super-resolution (SISR) method.
- To improve the quality of high-resolution image reconstruction from low-resolution inputs.
Main Methods:
- Adaptive fractional-order gradient interpolation and reconstruction.
- Construction of interpolated image gradient via optimal fractional-order gradient based on image similarity.
- Minimum energy function for final high-resolution image reconstruction.
Main Results:
- The proposed method effectively reconstructs high-resolution images with rich texture details.
- Maintains structural similarity even under large zoom conditions.
- Outperforms current single image super-resolution techniques in experimental evaluations.
Conclusions:
- Fractional-order gradient methods offer an additional degree of freedom for optimizing image quality.
- The presented adaptive fractional-order gradient approach significantly enhances SISR performance.