Depth-of-field expansion method based on multidimensional structure and edge-guided correction

Area of Science:

• Computer Vision
• Image Processing
• Digital Imaging

Background:

• Multi-focus image fusion aims to extend the depth of field, producing images with comprehensive focus.
• Effective pixel-level focusing detection and region optimization are critical for successful fusion.
• Existing methods may struggle with preserving details or introducing artifacts.

Purpose of the Study:

• To propose a novel multi-focus image fusion method, the multidimensional structure and edge-guided correction (MSEGC).
• To enhance the preservation of image details and non-texture regions.
• To reduce edge artifacts in fused images.

Main Methods:

• A redesigned pixel-level focusing evaluation function is employed.
• Edge-guided decision correction is utilized to mitigate artifacts.
• The MSEGC method integrates multidimensional structure analysis with edge guidance.

Main Results:

• The MSEGC method demonstrated significant improvements in objective evaluation metrics, ranging from 22% to 50% compared to other approaches.
• Verification using public datasets and semiconductor inspection images confirmed the method's effectiveness.
• The fused images exhibited enhanced visual quality and detail preservation.

Conclusions:

• The proposed MSEGC method offers superior performance in multi-focus image fusion.
• MSEGC effectively balances detail preservation with artifact suppression.
• This technique provides a valuable advancement for applications requiring high-quality, fully focused images.