Scanned image descreening with image redundancy and adaptive filtering

Area of Science:

• Digital Image Processing
• Computer Vision

Background:

• Electrophotographic printers utilize halftoning for continuous-tone images, leading to screen-like artifacts in scanned copies.
• Existing methods struggle to effectively remove these artifacts while preserving image quality.

Purpose of the Study:

• To propose a new model for scanned halftone images accounting for printing distortions and halftone patterns.
• To develop an adaptive filtering-based descreening method for high-quality continuous-tone image recovery.

Main Methods:

• A novel scanned halftone image model incorporating printing distortions and halftone patterns.
• An adaptive filtering approach utilizing screen frequency and local gradient features.
• An image redundancy-based denoising algorithm and an edge-preserving filter for enhanced sharpness.

Main Results:

• The proposed method successfully recovers high-quality continuous-tone images from scanned halftone images.
• Experimental results show superior performance compared to state-of-the-art methods.
• The method achieves very sharp edges and significantly cleaner smooth regions.

Conclusions:

• The developed adaptive filtering method effectively removes screen artifacts from scanned halftone images.
• The proposed technique offers a significant improvement in recovering high-quality continuous-tone images.
• This approach provides a robust solution for descreening scanned documents.