Face Mask Quality Control Inspection
Machine vision and deep learning ensure high-quality masks to protect frontline health workers
Graphical programming environment for deep learning-based industrial image analysis
There are different types of face masks that help prevent the spread of disease, from a typical surgical mask to an N95 respirator. This equipment is crucial to help keep workers safe. However, these masks only protect users if they are defect free and pass stringent ISO standards, which is challenging when demand far exceeds existing supply. The quality of masks should be rigorously checked to prevent defective products from making it to the market, including checking for flaws such as embedded hair or stains, measuring the width of the mask, checking for the presence of straps, and determining that the straps are correctly attached to the mask.
By leveraging both machine vision and deep learning technology manufacturers can ensure masks are produced in compliance with ISO standards and catch defective masks before they are shipped. Cognex In-Sight 8402 vision system detects the presence of facemask components such as earbands and strap welds, while also measuring the width of the masks to ensures they are manufactured to the correct size. Many defects, however, are difficult to predict and program with traditional machine vision algorithms. Cognex Deep Learning is trained with as few as 50 sample images to easily locate and classify random defects, such as rips, stains, and stitching errors.