Colorless image processing technique combining chromaticity transfer and image feature extraction in visual graphic...

Area of Science:

• Computer Vision
• Image Processing
• Computational Photography

Background:

• Traditional image processing methods suffer from color inconsistency, edge bleeding, and poor feature matching accuracy.
• Existing techniques often fail to achieve high-fidelity results in colorless image processing tasks.
• There is a need for robust and efficient frameworks for advanced visual graphic design applications.

Purpose of the Study:

• To present a verifiable framework for high-fidelity colorless image processing.
• To address limitations of traditional methods by introducing depth-guided chrominance transfer and multi-level feature extraction.
• To establish a new benchmark for colorless image processing in terms of accuracy, efficiency, and robustness.

Main Methods:

• Developed a dual-channel processing pipeline for RGB-D images.
• Introduced a weighted non-local Laplacian algorithm for cross-modal consistency.
• Implemented a multi-level feature hierarchy and a brightness-guided edge preservation mechanism.

Main Results:

• Reduced color reconstruction error by 60% compared to the Welsh method (0.020 ± 0.003).
• Achieved high edge fitting accuracy (0.93) with minimal over-segmentation (8%).
• Boosted matching recall to 91.4% ± 2.1% and significantly reduced semantic alignment error.
• Outperformed diffusion-based colorization by 3.7 dB PSNR with 37% faster processing speed (9.4 FPS).

Conclusions:

• The proposed framework offers a transparent, efficient, and robust solution for high-fidelity colorless image processing.
• Demonstrated significant improvements in accuracy and speed over existing methods on benchmark datasets.
• Establishes a new standard for visual graphic design applications requiring precise image manipulation.