Spectral properties inspired image enhancement for crystalline polymer powder impurity detection

Area of Science:

• Materials Science
• Polymer Chemistry
• Image Processing

Background:

• Polymer powder quality is assessed by contaminant levels, but detection is difficult due to poor image quality from production sites.
• Existing machine vision methods struggle to accurately identify impurities like decomposed polymers or foreign particles in crystalline polymer powder.
• Spectral properties of crystalline polymers offer a potential avenue for improved impurity detection.

Purpose of the Study:

• To develop an efficient and robust image-based method for detecting and quantifying impurities in crystalline polymer powder.
• To enhance image quality for better differentiation between normal polymer particles and contaminants.
• To improve the accuracy and reliability of polymer powder quality assessment.

Main Methods:

• A channel-weighted image enhancement approach inspired by spectral properties was designed to highlight impurity differences.
• An adaptive thresholding method, utilizing prior knowledge of powder attributes, was employed for impurity pixel classification.
• The algorithm was evaluated on a dataset of 119 high-resolution images from a chemical facility.

Main Results:

• The proposed channel-weighted image enhancement demonstrated superior selectivity for impurities compared to conventional methods.
• The adaptive thresholding effectively categorized impurity pixels based on enhanced image features.
• The method proved capable of detecting contaminants with an average size of 43 pixels in the evaluated dataset.

Conclusions:

• The developed image-based impurity detection method offers a precise and robust solution for crystalline polymer powder quality assessment.
• The spectral property-inspired image enhancement technique significantly improves the visibility and detection of contaminants.
• This approach addresses the limitations of machine vision in challenging production environments, enabling more reliable quality control.