Intelligent cameras with integrated processing units that perform image processing tasks on-device have been on the market for some time. But only resource-saving AI accelerators and machine learning methods enable completely new alternative operating and development concepts to emerge. This opens up the creation process of embedded vision applications to new user groups and thus opens up entirely new application areas.

The familiar development environment installed on a PC will be replaced by an AI Vision Studio where image processing is no longer programmed but trained with application images. Easy-to-use tools cover the entire development workflow from the first image to the finished application - without the need for special prior knowledge in deep learning, image processing or application programming. And last but not least, an app-based operating system ensures that intelligent cameras perform and switch vision tasks as flexibly as a smartphone.