Advances, Systems and Applications
Table 1 Summary of related work
From: Enhancement of damaged-image prediction based on digital twin technology
Reference | Model | Application | Key contributions |
|---|---|---|---|
Gillette and Bertozzi (2007) [6] | Cahn-Hilliard equation model | Binary image restoration | Added fidelity item based on CH equation for binary image restoration |
Burger (2009) [7] | \(TV-H^{-1}\) model | Grayscale image restoration | Removed free energy term in CH model for grayscale image restoration |
Kim (2019) [8] | CH equation | Image with broken stripes restoration | Introduced preprocessing method using anisotropic diffusion equation before solving CH equation |
Zou (2021) [9] | Perona-Malik equation and CH equation | Image restoration at boundary | Combined nonlinear diffusion coefficient and CH equation for smooth repair effect at the boundary |
Carrillo et al. (2021) [14] | CH model as preprocessor | Neural network image recognition | Improved recognition accuracy of damaged data using CH model as preprocessor |
This paper | Second-order numerical scheme based on the convex decomposition method | Image restoration | Proposed a time-based second-order numerical scheme for better restoration effect and improved prediction accuracy of the neural network model |