Software update with new features and trial month for AI training software

02-03-2022 | IDS | Test & Measurement

Users of the all-in-one embedded vision system IDS NXT ocean now have a complete range of new features at their disposal. These comprise multi-ROI ('Region Of Interest') for AI-based object detection and the opportunity to utilise different neural networks for different ROIs in one image through Vision App. Also, there are binning, line scan mode and performance and configuration improvements. New customers may also select whether they want to utilise the IDS NXT lighthouse training software on AWS – as previously provided by IDS – or now also through Microsoft Azure Cloud.

With Microsoft Azure Cloud Services, it now supports another hoster for the AI training software IDS NXT lighthouse in addition to AWS. The offer is targeted, particularly at companies that cannot work with AWS due to regulatory requirements. The scope of functions and services and the operation of the training software are the same with both hosters. Instead of setting up their own development environment, users may start training their own neural network straight away in IDS NXT lighthouse – even with no prior knowledge of Deep Learning or camera programming. This comprises three essential steps: Upload sample images, label them and start the fully automatic training. The resulting network is optimised for use with IDS NXT industrial cameras.

The company allows all customers to activate a complimentary trial month and extensively test IDS NXT lighthouse. No sales contact is needed to activate the full trial licence. All they have to do is register with their IDS user account at IDS NXT lighthouse and choose the appropriate option through the process. The licence key is automatically provided.


The current software release also heralds a regular cycle for function updates. In future, the company will release new features approximately every three to four months, which will improve the IDS NXT system still further and provide customers even more scope for designing their image processing applications.

By Natasha Shek