The industrial camera sector has evolved rapidly in recent years, with camera manufacturers focusing on three key aspects: "smaller, faster, and more intelligent". With the unstoppable progress in sensor technology and the increasing automation of production processes, new opportunities are constantly opening up to optimise the performance of industrial cameras or to better adapt them to requirements. These evolving steps towards more compact designs, higher speeds and intelligent functionalities have been instrumental in the further development of a wide range of industrial sectors. This technical article takes a look at the significance of these three camera classes and highlights how they support industrial image processing.
Tiny cameras for embedded applications
Single-board cameras can now be designed so small that they fit completely into the lens holder. Such tiny devices are therefore particularly suitable for small device construction and embedded vision applications. In this segment, IDS offers models with dimensions of only 29 x 29 x 7 mm. These small industrial cameras of the X-series are available in a wide range of housing variants for the most variable integration possible. From board level cameras with or without lens holder or housed in plastic to full zinc die-cast housings, they are cost-optimized for the most essential features. This makes them perfect for price-sensitive and high-volume applications. And thanks to their small size, these budget-friendly cameras can also be easily integrated into a wide variety of image processing systems.
Standard components for accessories and connections or interfaces ensure easy use. All common C-/CS-Mount as well as S-Mount lenses can be used and with Type-C connector and USB3 Vision standard they can be easily put into operation or integrated into a vision system without many accessories or complex development tools.
With this camera category, IDS addresses another market requirement. Many applications do not necessarily require cameras with high-end image sensors. The use of lower-priced variants that have been "downsized" by the sensor manufacturers only to a small extent in terms of resolution, frame rate and shutter technology, for example, compared to premium models, results in additional cost-saving potential here.
45 megapixels without any loss of speed
If cycle rates are to be increased, inspection times shortened and objects, such as wafers or panels, are to be analysed, measured or checked for defects with high precision from greater distances, large-format sensors are used. In this way, the required images or inspection points with the largest possible field of view can be reduced to a minimum and the throughput increased. In order to transmit the data-intensive camera images generated to an image processing unit in full resolution, at a high frame rate and with very low latencies, a high-speed interface is also required.
With the uEye Warp10, IDS has launched a camera family that transmits data particularly fast and effortlessly over long distances thanks to 10GigE in the Gigabit Ethernet-based network. Onsemi global shutter sensors with up to 45 MP are used as image sensors. In combination with the thread-based and thus industry-suitable TFL mount, these image sensors with up to 2" image diagonals can be exposed without edge shading.
Since such a high-speed system requires significantly more heat dissipation under full load than standard industrial cameras, a housing design with cooling ribs is used here. An additional active cooling plate even allows for a much more compact design than other 10GigE systems with passive coolers.
Image analyses (without PC) directly on the camera
You want to sort according to quality, categorise according to classes or detect objects or features and the image processing system should work in the smallest possible space, if possible even without an additional PC? In such cases, AI, edge computing and embedded vision provide more flexibility and productivity. Representatives of such compact embedded vision devices are intelligent cameras that not only capture images and send them to host PCs for further processing, but also perform the image analysis themselves with an integrated AI accelerator and generate process-relevant results completely independently.
These device classes thus provide data that cannot be generated with classic industrial cameras or only with additional system components - and thus take on a new role in connected systems. They are no longer just part of the image processing, but act as a complete image processing system. For communication, their interfaces support industrial protocols such as OPC UA, thus exchanging result data directly with machines and controllers and thus working seamlessly in Industry 4.0 applications - a milestone for the convergence of image processing and PLC. This creates groundbreaking synergies for the future.
The application decides!
So not all cameras are the same! Anyone talking about industrial cameras can have very different needs. Sensor resolution, data transfer speed and range of features are often not the decisive factors. The focus is therefore always on the application. Regardless of whether it is automation, quality assurance or self-driving vehicles - the requirements for suitable industrial cameras can already be specified in a more targeted manner by the characteristics smaller, faster, more intelligent.