SolVisionCase Study
AI Quality Control for Medical Masks Manufacturing
AI and machine vision mask defect inspection
Medical masks and epidemic prevention
Medical grade masks are capable of filtering bacteria, bodily fluids, airborne particles, and are the first line of protection against epidemics and viruses. Mask production involves processes such as material cutting, ear band welding, and packaging which are usually automated. Quality inspection however, still relies on manual execution and is labor-intensive to complete.
Large variations of masks and defects
Mask defects come in all shapes, sizes and appearance, from missing wires or components, incorrect stitching, ripped holes, stains, edge protrusions, to damaged layers. Masks also tend to be unsystematically distributed on the production line. Given the uncertainties and differences, it is difficult to recognize all types of potential defects using traditional inspection methods.
AI deep learning inspection solution
SolVision solves the problem of having too many defect variations, by allowing an image-based inspection system to be developed and labeling defects on sample images for the AI model to learn. After training, the system can recognize different defects on captured images so that faulty products can be identified and eliminated from the production line.
AI Inspection
Protrusion
Cut Mark
Missing Nose Clip
Dirt
Broken filter layer
Broken filter layer
Excess Nose Clip
Hole
Loose Strap
Dropped Strap
