Alleviating dirty-window effect in medium frame-rate binary video halftones.

Area of Science:

• Computer Vision
• Image Processing
• Display Technology

Background:

• Video display devices often use fewer bits per pixel than required, necessitating video quantization.
• Halftoning is a quantization method that can introduce artifacts, notably the dirty-window effect (DWE) in medium frame-rate binary videos.

Purpose of the Study:

• To propose and evaluate novel video halftone enhancement algorithms specifically designed to reduce the visibility of the dirty-window effect (DWE).
• To develop algorithms that minimize DWE while adhering to spatial quality constraints, with variations in complexity and temporal quality considerations.

Main Methods:

• Development of three distinct video halftone enhancement algorithms targeting DWE reduction.
• Assessment of algorithm performance using objective measures to quantify DWE in both original and enhanced halftone videos.
• Evaluation under varying constraints, including spatial quality, reduced complexity, and combined spatial and temporal quality.

Main Results:

• The proposed algorithms demonstrate a reduction in the dirty-window effect (DWE) in medium frame-rate binary video halftones.
• Algorithm variations offer trade-offs between DWE reduction, spatial quality preservation, and computational complexity.
• Successful mitigation of temporal artifacts under specified quality constraints.

Conclusions:

• The developed video halftone enhancement algorithms effectively reduce DWE, improving the visual fidelity of quantized video.
• The algorithms provide flexible solutions for DWE reduction, catering to different application requirements regarding quality and complexity.
• This work contributes practical methods for enhancing video display quality in resource-constrained environments.