news-blog.jpg

News

Headwall & perClass Partner to Advance Embedded Vision Solutions with Intuitive Spectral Imaging

Industry-leading platform paired with real-time onboard classification on processing/production lines

Bolton, MA, June 25, 2020 – Headwall Photonics, a world leader in hyperspectral imaging platforms and instrumentation, and perClass BV, provider of innovative and reliable software for production deployment of machine learning and spectral imaging solutions, announced a partnership today that will help meet challenging needs for customers worldwide.

“Markets such as food processing, high-tech manufacturing, and material-sorting for recycling now have a much easier path to exploit the power of hyperspectral imaging,” says Christian Felsheim, Director of Headwall Photonics Europe. “Headwall’s new Hyperspec® MV.X embedded hyperspectral imaging platform is designed from the start to run spectral classification algorithms such as those developed by perClass Mira® and output actionable data in real time, enabling automated sorting systems to take immediate action along the line.”

Dr. Pavel Paclik, Founder and Managing Director of perClass BV, agrees and adds, “Our mission at perClass is to significantly lower the barrier to entry for the vast number of industries that can benefit from an AI-based workflow that creates highly accurate models without the need for cumbersome coding. We’re excited to be working with Headwall’s architecture so that customers can fully benefit from the many advantages our own automated machine-learning platform offers.”

Accurate and reliable classification models can be generated using the Headwall MV.X and perClass Mira software on a PC. Those models are then uploaded to IP67-rated MV.X imaging systems on the production line with no need for dedicated computers to store and process the raw hyperspectral data. Amazingly the compact MV.X systems host a runtime version of perClass Mira that outputs the classified data through an industry-standard GenICam interface to robotic systems to take action along the line without any further processing.

The pairing of Headwall’s MV.X and perClass’ Mira platforms allow customers to rapidly deploy solutions that can be easily updated if requirements or situations change; for example, the material being inspected changes or even more accurate classification models are developed over time.

About Headwall

Headwall is a leading designer and manufacturer of complete spectral instrumentation solutions for remote sensing, advanced machine vision, and government/defense markets. With a worldwide base of end-user and OEM customers, Headwall enjoys a market leadership position through the design and manufacture of spectral solutions that are customized for application-specific performance. The Company has three European locations in Belgium, Germany, and Italy. Hyperspec is a registered trademark of Headwall Photonics. European headquarters operations at Headwall BVBA are located near Brussels, Belgium. For more information, visit www.headwallphotonics.com.

About perClass

Since 2008, perClass has been providing software for design and production deployment of machine learning and spectral imaging solutions. It is used in the most demanding applications, such as food quality control, industrial sorting, medical diagnostics or traffic accident detection.

The perClass Mira user-interface radically simplifies interpretation of spectral images. Within minutes users can create sophisticated classification solutions without machine learning expertise or programming. perClass runtime technology enables GPU accelerated real-time deployment in custom applications in both PC and edge computing scenarios. perClass and perClass Mira are registered trademarks of perClass BV. Find out more at www.perclass.com.

For more information, please contact:

Ross Nakatsuji, Headwall Marketing Communications

580 Main Street

Bolton, Massachusetts 01740

+1-978-353-4051

rnakatsuji@headwallphotonics.com

###

Link to original Press Release

Subscribe Now

Recent Posts

Posts by Topic

see all