Display

Adaptive processes through cameras
with AI

This changes the way in which machines are planned and systems are implemented. Instead of following pre-programmed, fully controlled tasks, these cameras make decisions autonomously. This means more flexibility in variable environmental conditions - and enables a robot to precisely recognize chaotically placed parts, grip them in the correct position and process them further. This allows complex and variable tasks to be managed effectively.

From object localization and OK/NOK inspection to anomaly detection and more: depending on the AI method and purpose, intelligent cameras such as IDS NXT provide the right information. © IDS Imaging Development Systems GmbH

Everything in view even at high clock rates

Hardly any other component collects and interprets as much data as image processing. Whether object localization, segmentation or anomaly detection: intelligent cameras determine the desired information in a targeted manner. In contrast to rule-based image processing, relevant characteristics (such as good/defective) do not have to be described manually using code, as the AI learns these independently based on training images. This also makes it possible to implement applications that were previously unprofitable due to the disproportionately high time and cost involved, for example with small batch sizes or objects that vary greatly or unpredictably.

Different options, common goal

IDS Imaging Development Systems GmbH will be presenting the IDS NXT complete AI system at automatica. It supports customers from image acquisition to the execution of neural networks - and is characterized by user-friendly workflows. The system is continuously being developed further. For example, IDS NXT malibu, a new camera prototype that significantly accelerates inference, will be on display at the trade fair stand. Perfect for getting started: the IDS NXT ocean Creative Kit.

Advertisement

IDS offers DENKnet, a high-performance no-code solution for AI-based image analysis for those who want to run AI on a PC rather than a camera. The training of a neural network can be started with as few as 15 images.

What both systems have in common is that they want to make the use of AI vision as easy as possible for users. Find out more: directly at IDS, automatica, B5.203.

  • Xing Icon
  • LinkedIn Icon
Advertisement
Advertisement

You might also be interested in

Advertisement
Advertisement
Advertisement
Advertisement
Advertisement
Advertisement
Advertisement
Subscribe to our newsletter
Advertisement
Back to home