Skip to content

Embedded industrial camera technologies

Linux based open source cameras designs

Trusted camera provider for

Our embedded camera platforms

The cameras are modular by nature and can be tailored to fit client-specific needs

qtec C-Series

APU

AMD Ryzen Embedded V1605B w/ Radeon Vega 8 Graphics V1605B

qtecOS

GNU/Linux (OpenEmbedded/Yocto)

I/O

CFast, ETH (x2), USB 2.0 (x3), Mini DP, WiFi, Trigger, Flash & GPIO

From € 3.200*

See datasheet

qtec T-Series

APU

Qualcomm Snapdragon 820

qtecOS

GNU/Linux (OpenEmbedded/Yocto)

I/O

Flash, ETH, USB 3.0, USB 2.0, Micro HDMI, WiFi, Trigger, Flash & GPIO

From € 1.500*

See datasheet

qtec M-Series

APU

Qualcomm Snapdragon 820

qtecOS

GNU/Linux (OpenEmbedded/Yocto)

I/O

Flash, ETH, USB 3.0, USB 2.0, Micro HDMI, WiFi, Trigger, Flash & GPIO

From € 1.300*

See datasheet

*The prices above are indicative and subject to change depending on sensor choice and modification needs.

Modular camera architecture

Exchangeable camera heads
Embedded processing
Open source

1/3

Exchangeable camera heads

Our exchangeable camera heads can be utilised on different processing platforms, which allow our clients to tailor camera heads for specific applications. This facilitates the testing of alternative solutions for various projects and scientific research projects. It also enables the replacement of a camera body without having to replace the camera head once the camera hardware becomes obsolete. The architecture is modular by nature and comprises a growing number of heads into which a variety of CMOS, CCD, InGaAs, and microbolometer sensors can be mounted with various configurations and lens interfaces.

2/3

Camera body with embedded processing

One of our most popular camera platforms is the QT5022 platform with AMD APU. The QT5022 camera body contains two main computing units, an FPGA and AMD (APU), which features X86 CPU cores and Radeon Graphics GPU in a SOC. The FPGA in the camera controls the settings of the sensor and interfaces to the different sensors. Several operations are performed by the FPGA, such as image correction, white balance correction, perspective correction, and illumination correction, thus offloading the main CPUs of these tasks. The preprocessed images are transferred over a PCI Express interface to the AMD APU. Where further computing power is required, an additional GPU can be added within the body of the camera.

3/3

Non-proprietary application software

Qtechnology is a big player in the Open Source embedded vision ecosystem with contributions in all the stack and a synonym of innovation and openness. Even if you are not familiar with our industrial vision solutions, you are most likely already using our software in your Phone, Browser or Cloud Provider. More than 1000 patches on the Linux kernel, GStreamer and OpenCV, among others, probe our devotion to Open Source. Our industrial cameras run the latest upstream software, tailored to their performance, so you simply have to focus on your area of expertise. No need to learn new libraries or fears of getting locked in a closed ecosystem. The same software that can run on your notebook and make use of your webcam, can run on all our highly specialised industrial cameras.

Artificial Intelligence

Neural Network cameras

Large variations in environment and/or products can sometimes make it difficult to
extract meaningful patterns from images.

Some tasks require analysing very large amounts of data, or contain complex features which are difficult to identify. For cases where traditional image processing falls short, we can exploit neural networks. Common applications include advanced classification, detection, tracking or segmentation.

Qtechnology houses state-of-the-art computing facilities for designing and training the right neural network architecture, ready for deployment on our cameras.

Technology

Advanced embedded camera technologies

With a modular camera designed for easy adoption to different sensors, we can implement specific sensor mechanics, lens mounts, lighting, and more, in a customised camera head that can easily be mounted on our camera bodies.

Cameras can be customised based on client-specific needs regarding mechanics, sensors, housing, working environment, etc. This includes tailored head configurations such as angled sensors, special optical mounts, filters, and built-in LED light.

We are also able to provide specific software interfaces, protocols, processing algorithms, pipelines, etc.

Optical fields

See beyond the visible with Qtechnology cameras with imaging technology for different optical fields

Infrared
Visible light
Multi- & Hyperspectral
X-ray

IR cameras for NIR, SWIR & LWIR imaging

Qtechnology cameras can be used for applications based on different inferred imaging technologies beneath the visible colour spectrum.

NIR (650 – 950 nm)
By replacing infrared filtering in the cameras while using NIR enhanced image sensors, our cameras can be used for near-Infrared Imaging imaging.



SWIR (900 – 1700 nm)
With special InGaAs sensors, the Near-infrared imaging spectrum can be increased to cover up to 1700 nm. Using InGaAs sensors, it is further possible to use our cameras for hyperspectral imaging, spectroscopy, biological imaging, water/moisture detection, industrial sorting, Internal defects, etc.

LWIR (8 – 14 um)
LWIR applies uncooled microbolometer sensors, which can be used for industrial thermal imaging such as building inspection, industrial insurance inspection, industrial process control, etc.

RGB cameras for visible light imaging

Visible light imaging technology is the majority of industrial machine vision applications. In particular, in applications where colour is an important characteristic. However, many applications require very specific light wavelengths to emphasise certain features of the object of interest.

If the colour of the feature of interest is known, using a similar wavelength of colour LED will highlight this specific feature. For instance, green light makes green features appear brighter while red light will make green features appear darker.

At Qtechnology, we can optimise vision systems by customising different coloured LED modules or combinations specifically to each client application

HSi & RGB-Ir cameras for multispectral & hyperspectral Imaging

Some vision solution requires our multi-spectral or hyperspectral imaging technology.

Multispectral imaging is a combination of a few bands of the spectra at the pixel level. An example is our 3 CCD camera where each colour spectra is optically separated in red, green and blue and recorded at pixel level on three CCDs one per colour.”

With the use of low noise sensors, hyperspectral image cubes can also be captured using spectrographs or specialised sensors with narrow-band spectral filters at the pixel level.

These cubes can be used for spectroscopy to identify spectral features and signatures of different materials. It is not only used to separate specific materials but also to make qualitative measurements of the analysed objects.

Cameras for x-ray imaging

By combining our low noise image sensors with scintillators, we also provide cameras that can be used for x-ray imaging.

This technology can, for instance, be used for the detection of metal, internal damages, hidden components, cracks, welding flaws, and determination of dimensional measurement.

Camera configurations

V4L2 API Controls

Qtechnology cameras are easily configured through the web interface and images can be checked and downloaded through a standard web browser, without installation of any special programs or drivers.

Through the web interface, it is possible to control all the camera features such as image formats, image size, regions of interest, frame rate, exposure time, triggers, gain, etc. These features are controlled through V4L2 API controls, which are also available controls from user applications.
Other special features built into the web interface are auto white balancing, focus adjustment measures, histograms, sensor calibration, etc.
All configurations can be saved and restored for use in user applications or for setting up other cameras.

Cameras can be customised to your needs, if it’s different sensors, mechanics, environment, etc. We are also able to supply specific software interfaces, protocols, processing algorithms, pipelines, etc.

Application

Software stack & hardware acceleration

There are multiple ways of developing your application. Qtechnology’s industrial cameras give you complete freedom to choose the software stack that solves your computer vision problem. There is no need to build your applications on top of a proprietary library or application.

All the interaction with the camera and the FPGA is done through the standard Video4Linux API. This includes not only the different format converters but also the programming of the advanced components such as the LUT, camera calibration or Illumination Correction.

The access to the hardware acceleration is also done with a standard API, in this case, OpenCL, allowing your application to be ported seamlessly on all the different Qtechnology cameras, even on different architectures.
Nevertheless, we also support other parallel computing platforms such as CUDA on our NVIDIA-based industrial cameras.