Adaptive Vision Software Success Stories
Max Reijngoudt
Adaptive Vision software helps machine vision engineers build image-processing applications for automation and quality inspection. The platform is still commonly searched as Adaptive Vision, and today it’s available from Zebra Technologies as Zebra Aurora Vision. This page collects success stories that show the software in action—from dimensional measurement to PCB positioning.
A practical way to build vision applications with Adaptive Vision
Adaptive Vision software is based on a dataflow-based approach, including a variety of pre-built image analysis filters. Tailor-made for professionals, it simplifies the creation of common applications and streamlines the development of customized projects. This adaptability makes it a go-to tool for computer vision engineers addressing diverse quality inspection and automation needs in the industrial sector.
Adaptive Vision software toolkit
The Adaptive Vision software webpage showcases a range of computer vision software packages for image processing including:
- Aurora Vision Studio: A powerhouse of simplicity, offering a graphical low-level programming environment for quick development and easy customization by machine vision engineers.
- Aurora Vision Library: Ready-to-use features for C++ and .NET programming, giving you the freedom to develop your own software for even the most complex applications.
- Aurora Vision Deep Learning: This Aurora Vision software add-on allows you to select images, mark defects/labels, and train the software for your specific application. Dive deep into optical character recognition with the pre-trained Deep Learning OCR tool. This tool is perfect for deciphering complicated, damaged, or blurred characters. It's the ultimate solution for overcoming visual challenges in character recognition.
Using Adaptive Vision for Airport tray verification
In this application we demonstrate show how to use Adaptive vision for an airport tray verification (2D classification + 3D measurement). Adaptive Vision software by Zebra Technologies includes advanced deep learning capabilities, allowing users to create customized algorithms tailored to their specific needs. One notable example of this flexibility is a project carried out by the Technical University of Vienna, specifically the Institute for Production Engineering and Photonic Technologies. The development at the Institute was led by a team of experts, Dipl.-Ing. Fabian Singer and Dipl.-Ing. Laurenz Pickel, who specialize in providing bespoke machine vision solutions. In this project, the team developed a dual-function system that combines both 2D image analysis using deep learning and 3D measurement using classical machine learning algorithms.
Deep Learning add-on for Adaptive Vision Studio, brings a major step forward for machine vision applications. It offers a set of ready-to-use tools that you train with good and bad samples, enabling automatic detection of defects or features. Behind the scenes, it leverages large neural-network models optimized by a research team for industrial inspection. For users, these capabilities are delivered as simple filters with only a few parameters, supported by intuitive graphical tools that make the training process easy to run and manage.
Deep learning model
The first component of the system is a deep learning model, designed to analyze images of airport security trays and classify them into one of three categories:
- Empty trays
- Trays with objects left behind
- Trays requiring attention due to liquid spills
This algorithm was trained on a dataset of 350 images, ensuring accurate detection and classification to improve airport security measures. The deep learning approach enables the system to provide real-time feedback and detailed visualizations, such as heatmaps, that guide security personnel in identifying issues swiftly.
3D Measurement with Classical Machine Learning
The second component of the system focuses on 3D analysis. Unlike the 2D image classification, this part of the system uses classical machine learning algorithms to perform precise measurements of the tray and its contents:
- Measures the entire tray volume
- Calculates the volumes of the objects inside the tray
- Determines the total volume of the tray’s contents
- Identifies the maximum fill height
This volumetric analysis provides a comprehensive overview of the tray's status, allowing for efficient resource management at security checkpoints.
Integration and Efficiency with Adaptive Vision
The combination of deep learning for 2D classification and classical machine learning for 3D measurements results in a highly effective solution for airport tray verification. The system operates in real-time and provides security staff with accurate insights, enhancing the precision and speed of airport security procedures. Adaptive Vision software, in conjunction with machine vision cameras, significantly raises the precision and efficiency of airport tray control, ensuring enhanced security protocols at checkpoints. For inquiries regarding custom machine vision solutions, we invite you to reach out directly to the Institute or contact the lead scientists of this project. You can find more information at TU Wien - Institute for Production Engineering and Photonic Technologies.
Adaptive Vision for Diameter inspection
Diameter Inspection Overview
For this vision project, engineers used Adaptive Vision Software to develop an application dedicated to diameter dimension inspection of tempered steel rings, including roundness and inner diameter verification. By harnessing machine vision cameras, this Adaptive Vision Software workflow provides highly accurate assessment—supporting inspection tolerances of up to ±0.1 mm.
The challenge
Tempered steel rings can be difficult to measure quickly and consistently in production. The team needed a solution built with Adaptive Vision Software that could:
- Verify inner diameter and roundness reliably
- Deliver consistent results for industrial environments
- Provide fast, repeatable measurement output Support strict acceptance criteria (up to ±0.1 mm)
The solution - Adaptive Vision Studio
Using Adaptive Vision Studio, the engineers created a robust inspection workflow powered by machine vision cameras. The application was designed to:
- Acquire images from machine vision cameras
- Apply image processing to isolate ring geometry
- Measure inner diameter and evaluate roundness
- Output results in real time to support quality inspection and industrial automation
In this success story, Adaptive Vision Software plays a pivotal role by providing a platform for diameter inspection that goes beyond conventional limits. Its adaptable approach—integrated with machine vision cameras—supports real-time measurements and analysis, ensuring the inner diameter of steel rings meets stringent quality standards.
The results - Accurate inspection
With the inspection workflow built in Adaptive Vision Software, the team achieved:
- Accurate inner diameter inspection aligned with quality requirements
- Reliable roundness verification for critical parts
- Real-time measurement and analysis suitable for production workflows
- Inspection precision enabling tolerances up to ±0.1 mm
This project highlights the value of Adaptive Vision Software in advancing industrial inspection, improving efficiency while maintaining high accuracy for demanding dimensional measurements.
PCB positioning with adaptive vision software
PCB Positioning Overview
In this application, engineers used Adaptive Vision Software, particularly the advanced Adaptive Vision Studio, to create a high-precision solution for PCB positioning. The vision workflow was combined with a 5MP USB3 camera (such as the MER2-503-36U3M) to achieve accurate PCB position determination with measurement accuracy of up to 0.05 mm.
The challenge
PCB positioning demands reliable, repeatable alignment. The team needed a workflow built with adaptive vision software that could:
- Accurately determine PCB position with high precision
- Support customizable inspection logic for different PCB designs
- Integrate smoothly into existing operating software
- Maintain stable performance in real-world production environments
The solution - Adaptive Vision Studio Software
Using Adaptive Vision Studio as the core development environment, the team built a PCB positioning inspection workflow with Adaptive Vision Software and a 5MP USB3 machine vision camera. The solution emphasized:
- Precise positioning analysis for accurate PCB examination
- Flexible configuration through macro filters, enabling customized adjustments per application
- Fast integration via DLL export, allowing the positioning logic to be embedded into the customer’s own operating software
The results - Accurate positioning
With the PCB positioning workflow developed in Adaptive Vision Software, the team achieved:
- Accurate PCB positioning with measurement accuracy up to 0.05 mm
- A configurable inspection approach using macro filters
- Streamlined integration using DLL export for deployment
- A scalable software workflow suitable for ongoing production needs
This project demonstrates how Adaptive Vision Software can enhance PCB inspection and positioning by combining precision measurement, flexible workflow design, and integration-ready outputs.
