Stuttgart 2021

MICHAEL WHEELER, EDITOR-IN-CHIEF michael.wheeler@photonics.com

The autumn issue of Vision Spectra has been thought of as the “VISION issue,” set to coincide with the trade fair held every two years in November in Stuttgart, Germany. But because of COVID-19, Messe Stuttgart elected to call off the event this year, citing overwhelming response from the exhibitors.

With no virtual alternative, I resigned myself to the fact it would be another two years before we would have the chance to meet with exhibitors and attendees from more than 30 countries to see new products and learn about the profound impact vision is having on global industries.

Welcome news came this week when the show organizers announced VISION would resume next October and follow an “odd year” schedule going forward. Cue the conference’s Queen anthem — next October can’t come soon enough!

Meanwhile, online events are filling the void, with the Embedded Vision Summit that took place in September, and in November, A3’s International Vision Solutions Conference will feature training sessions discussing how vision systems can improve throughput and reduce defects.

Speaking of reducing defects, this month’s cover story examines the unique challenges posed in inspecting medical devices as diverse as catheters and artificial joints. Specifications are often tight, and defects such as pits or bumps barely the width of a human hair can be tough to spot and categorize, requiring a system with resolution of 25 µm — sometimes even 10 µm — per pixel. Be sure to read Hank Hogan’s “Vision Systems Perform Vital Checks on Medical Devices”.

Deep learning can help optimize defect detection as an alternative to “rule-based” systems. As we learn in Mario Bohnacker’s “Optimal Defect Detection with Deep Learning” , the drawback to previous deep learning detection methods was the large amount of training data required. But an innovative technology called “anomaly detection” requires far fewer training images — often between 20 and 100.

We round out this edition with “Hyperspectral Imaging Discerns Authenticity ofArtwork” from TruTag Technologies. And Ralf Muenster’s “CoherentSensing Holds Promise for Machine Vision” on page 40 examines the future potential for frequency-modulated continuous-wave sensing for 3D vision.

Enjoy the issue!
>