Accurate Detection for Zirconium Sheet Surface Scratches Based on Visible Light Images

Area of Science:

• Materials Science
• Surface Engineering
• Non-Destructive Testing

Background:

• Zirconium sheets are vital in chemistry and aerospace.
• Surface scratches degrade performance, impacting working life and fatigue resistance.
• Detecting these defects is challenging due to noise and complex textures.

Purpose of the Study:

• To develop an adaptive framework for detecting scratches on zirconium sheet surfaces.
• To address challenges posed by scattered noise and interlaced textures in scratch imaging.
• To enhance the accuracy and robustness of scratch defect detection.

Main Methods:

• Implemented an optimized dual-tree complex wavelet transform for noise removal.
• Employed an optimized relative total variation enhancement model for texture suppression.
• Integrated a novel framework with connection algorithms, replacing the Gaussian filter in Canny edge detection.

Main Results:

• The proposed framework effectively removes scattered noise.
• Interlaced textures are significantly suppressed.
• Experimental results demonstrate higher accuracy in scratch detection compared to standard methods.

Conclusions:

• The developed framework provides a robust and accurate method for detecting scratches on zirconium sheets.
• This advancement is critical for maintaining the performance and reliability of zirconium components.
• The approach offers a significant improvement in non-destructive evaluation of material surfaces.