Create innovative solutions with our state-of-the-art deep learning and vision knowledge. We specialize in developing custom, high-performance machine vision software integrated seamlessly with high-tech equipment and robotic control systems.
Combining technical know-how with machine learning
Nobleo is involved in developing machine learning solutions for high-tech systems and autonomous robotics. It is our ambition to apply machine learning by leveraging know-how of mathematics, physics, optics, electronics, and mechatronics to deliver innovative solutions for our customers.
Depending on the customer request, we can deliver an algorithm, a stand-alone application and/or a turnkey solution with integrated sensors and image processing. To co-develop machine learning algorithms efficiently we developed our own machine learning framework. The framework facilitates web-based data annotation, training & evaluation and deployment of machine learning networks on edge processing devices (see image).
To improve algorithm performance, we have developed a secure data storage solution to manage IP sensitive data. The data store is directly linked to the machine learning framework, providing easy access for data annotation and re-training of machine learning algorithms on data collected from devices “in-the-field”.
Key technical capabilities concerning Image processing & Machine learning
Object detection & tracking
Real-time neural network-based identification and tracking.
Object classification
Custom dataset training for industry-specific recognition.
Environmental understanding
Depth perception and scene analysis for autonomous navigation.
Sensor fusion for vision
Multi-sensor integration for enhanced perception.
Vision-Control Integration
Precise robotic actions based on real-time visuals.
Neural Network Training
Custom AI models for high-accuracy applications.
Multi-Spectral Imaging
Imaging techniques for advanced material recognition.
Machine Learning for Complex Control Framework
Our Machine Learning for Control framework creates advanced controller policies using simulation-based learning, tackling complex, non-linear systems that traditional methods struggle with.
Object detection & tracking
Real-time neural network-based identification and tracking.
Object classification
Custom dataset training for industry-specific recognition.
Environmental understanding
Depth perception and scene analysis for autonomous navigation.
Sensor fusion for vision
Multi-sensor integration for enhanced perception.
Vision-Control Integration
Precise robotic actions based on real-time visuals.
Neural Network Training
Custom AI models for high-accuracy applications.
Multi-Spectral Imaging
Imaging techniques for advanced material recognition.
Machine Learning for Complex Control Framework
Our Machine Learning for Control framework creates advanced controller policies using simulation-based learning, tackling complex, non-linear systems that traditional methods struggle with.
Assets
NoblAI Framework
Our NoblAI framework is a powerful platform for managing large datasets next to annotating, training, validating and deploying machine vision algorithms tailored to your needs.
Industries we service with our image processing & machine learning expertise
- Industrial systems & automation | Photonic wafer inspection
- Industrial cleaning | Robotized cleaning based on 3D imaging
- Offshore | Sensor fusion & tracking
- Recycling | Waste detection and recycling of refrigerators
- Agri & Food | orchard mapping and analysis
Specialized image processing & machine learning capabilities
- Quality inspection for high-mix low volume production | flexible tooling to support high-skilled staff making complex products in low volumes
- High-speed surface inspection with machine learning | leverage GPU power to run machine learning for inline applications
- Robotics integration | integrate machine learning in robotic systems for vision-based control
- In-house tools for incremental algorithm development | deployment and life-cycle management for machine learning algorithms
Contact us for more information about our Smart Robot Solutions
"*" indicates required fields
Can I help you?
Trusted by industry leaders
Frequently Asked Questions
What image processing and machine learning challenges can Nobleo Technology help solve?
We help with image processing and machine learning challenges in high-tech systems and autonomous robotics. This includes object detection and tracking, object classification, environmental understanding, sensor fusion for vision and vision-control integration. Depending on the customer request, we can deliver an algorithm, a stand-alone application or a turnkey solution with integrated sensors and image processing.
When is machine vision a better solution than traditional image processing?
Machine learning is useful when image processing tasks involve variation, complexity or recognition challenges that are difficult to capture with fixed rules alone. Examples include object classification, high-speed surface inspection, waste detection, orchard mapping and vision-based control. The right approach depends on the application, available data and required performance.
How does Nobleo Technology develop machine learning models for industrial applications?
We develop machine learning models by combining domain knowledge with data annotation, training, evaluation and deployment. Our NoblAI framework supports this process and helps manage machine vision algorithms tailored to specific customer needs. This enables incremental development and improvement of algorithms for industrial and robotic applications.
How does Nobleo Technology improve machine learning algorithms after deployment?
Algorithm performance can be improved by collecting relevant data from devices in the field and using it for annotation and retraining. An evaluation of the improved algorithm’s performance is caried out and reported to the customer, who can decide to update the deployment.
Can Nobleo Technology deliver more than just machine learning algorithms?
Yes. Depending on the request, we can deliver an algorithm, a stand-alone application or a turnkey solution with integrated sensors and classical image processing. This means we can support projects ranging from machine learning model development to complete machine vision solutions integrated with high-tech equipment or robotic control systems.
How do image processing and machine learning improve the performance of high-tech systems?
Image processing and machine learning can help high-tech systems recognise objects, detect defects and support more accurate control decisions. This can improve inspection, automation, robotic control and decision-making in applications such as photonic wafer inspection, industrial cleaning, offshore tracking, recycling and agri & food.