How to find anomalies with machine vision using IDS NXT
Learn how to detect anomalies in production processes using IDS NXT cameras and AI vision technology. This video demonstrates a complete workflow for anomaly detection – from capturing images and training a neural network to deploying a vision app on an IDS NXT camera.
You’ll discover why anomaly detection is essential when defects are unpredictable and cannot be classified easily, and how IDS NXT simplifies this task with intuitive tools like IDS NXT lighthouse and block-based app creation. Whether you need to identify subtle defects or automate quality checks, this guide shows you how to achieve accurate results quickly.
Topics covered
- Why anomaly detection is crucial for unpredictable defects
- Step-by-step workflow from image capture to AI training
- Using IDS NXT lighthouse for cloud-based CNN (Convolutional Neural Network) training
- Building vision apps with the block-based editor
- How to fine-tune anomaly thresholds for real-world accuracy
- Deploying and testing AI vision apps on IDS NXT cameras
Video timeline
- 00:00 - Topic introduction
- 01:04 - Machine vision use case
- 02:28 - Capture training images
- 03:27 - Prepare datasets for AI training
- 05:37 - CNN training
- 06:46 - Build a vision app
- 08:55 - Test & evaluate trained CNN
- 09:58 - Complete & test the vision app
- 11:02 - Deploy & execute on camera
- 12:29 - Summary
Who should watch
This video is ideal for quality engineers, machine vision specialists, and automation professionals looking to implement AI-based anomaly detection in manufacturing or electronics production.
Watch now to learn how to build and deploy anomaly detection with IDS NXT.
