Industrial x-ray image enhancement network based on a ray scattering model

Area of Science:

• Medical Imaging
• Image Processing
• Computer Vision

Background:

• X-ray images often suffer from low contrast and loss of edge details due to object complexity and ray scattering.
• Traditional enhancement methods struggle to preserve crucial image features.

Purpose of the Study:

• To develop an advanced method for significantly enhancing the quality of X-ray images.
• To address the limitations of existing techniques in preserving fine details and contrast.

Main Methods:

• A ray scattering model was employed to generate initial clear images.
• An encoder-decoder network with multi-scale feature extraction was utilized to add detailed information.
• A dual attention module and learnable UnsharpMasking were incorporated to selectively emphasize important features.

Main Results:

• The proposed method demonstrated a significant improvement in X-ray image quality.
• Experimental results confirmed the effective restoration of lost edge features and contrast.
• The technique successfully enhanced multi-scale and detailed information in the images.

Conclusions:

• The developed method offers a robust solution for enhancing degraded X-ray images.
• This approach holds potential for improving diagnostic accuracy in medical imaging.
• The integration of scattering models and attention mechanisms advances X-ray image enhancement.