Author

Lucas Steinmann

Lucas is the CTO of preML and responsible for the technical devlopment inside preML. I'm enthusastic about anything related to computer vision.

The European Machine Vision Forum (EMVF) is a remarkable platform for machine vision enthusiasts in Europe, bridging the gap between researchers and industry experts. This year, over 100 participants converged at the 6th European Machine Vision Forum in Wageningen, held on October 12th and 13th.

 

Image 1: Researchs and Industry experts meet at EMVF Wageningen 2023.

Real-world Machine Vision Challenges – Coping with Variability and Uncontrolled Environments

This year, Machine Learning took center stage at the European Machine Vision Forum, focusing on particularly unstructured environments like agriculture and industrial production. Machine learning methods have proven invaluable in coping with variability, but they come with prerequisites, most notably: suitable training data. The trend to tackle this challenge by aligning data or generating synthetic data with the aid of artificial intelligence or 3D rendering software can also be seen on the EMVF by the numerous talks which cover these topics. We submitted an extended abstract about one of our real world cases. The abstract was selected to be presented and I was happy to drive to Wageningen to present the project.

Our contribution

In a nutshell, we presented an AI-based CAD model alignment technique using keypoint regression for surface rectification in machine vision inspection systems. The method offers a practical solution for accurate alignment and rectification, allowing for precise quality control and quick adaptation to new variations in various manufacturing and industrial applications. The ability to automate the alignment process using deep learning significantly enhances the efficiency and accuracy of machine vision inspection systems, paving the way for advancements in automated quality assurance. Further, it shows an efficient way of coping with the challenge of variability and uncontrolled environments when implementing machine vision systems in the real-world.

About the European Machine Vision Forum

The European Machine Vision Forum, an annual event by the European Machine Vision Association (EMVA), fosters collaboration between the machine vision industry and academic research. It serves as a platform to share the latest research findings, address challenges in applications, explore emerging fields, and discuss potential research collaborations between industry and academia.

The forum’s primary goal is to expedite innovation by facilitating the rapid translation of new research into practical applications. It welcomes scientists, development engineers, software and hardware experts, and programmers from both research and industry sectors.

Image 2: By the way – we are one of the newest members at EMVA!

Big thanks to EMVA for organizing this event!

Cheers!

Lucas

Contact us at contact@preml.io for more information

Autor

Lucas Steinmann

Lucas is the CTO of preML and responsible for the technical devlopment inside preML. I'm enthusastic about anything related to computer vision.