Paperboard Coating Detection Based on Full-Stokes Imaging Polarimetry

Area of Science:

• Materials Science
• Optical Engineering
• Food Packaging Technology

Background:

• Current quality control for polyethylene (PE)-coated paperboard is manual, destructive, and offline.
• Defects like cracks and pinholes in PE coatings impact food packaging functionality.
• Effective inspection requires distinguishing between the PE coating and the paperboard substrate.

Purpose of the Study:

• To develop and validate an active full-Stokes imaging polarimetry system for classifying PE-coated paperboard.
• To enable non-destructive, in-line defect detection in the food packaging industry.
• To improve the quality and functionality assessment of extruded PE paperboard.

Main Methods:

• Utilized vertically polarized illumination and a novel full-Stokes imaging polarimetry camera.
• Implemented feature selection using distance correlation statistical methods.
• Employed a support vector machine algorithm with a nonlinear Gaussian kernel for classification.

Main Results:

• Achieved a classification accuracy of 99.74% for PE-coated paperboard and its substrate.
• Demonstrated the system's ability to discriminate between coating and substrate materials.
• Validated the polarimetry system on industrially manufactured samples.

Conclusions:

• Active full-Stokes imaging polarimetry offers a highly accurate method for material classification.
• The developed system can be integrated for in-process control in coated paperboard manufacturing.
• This technology enhances quality assessment for food packaging applications, reducing manual inspection needs.