Graphic of a network

Advantages of image processing networks

Ethernet as an infrastructural basis offers the advantage of a large possible distance (100m) between the camera and the computer processing the images. But even with smart cameras that have their own integrated image processing, Ethernet is the ideal interface for communicating the result data to the local network. With a transmission rate of 1 gigabit/s, the technology is fast enough for a wide range of application scenarios, including complex multi-camera applications with the necessary synchronisation mechanisms. The low-effort installation via standard network cable and the resulting fast, cost-effective commissioning are further arguments in favour of image processing via Ethernet.

Technical basics

Technically, image processing via Ethernet is based on standard Gigabit Ethernet components. In principle, it is also possible to use Fast Ethernet networks if the performance is sufficient for the application. As soon as several or even many cameras are used, high-performance switches with 10G ports can be useful for the congestion-free transport of data in the backbone. The transmission in the network often takes place in "jumbo frames", data packets with a maximum transmission unit (MTU) of 9000 bytes, instead of the 1500 bytes in an Ethernet standard frame. Jumbo frames noticeably relieve the CPU of the connected PCs. If the camera is not only to transmit data via the Ethernet cable, but also to draw power as a PoE Powered Device (PD), this requires the use of PoE switches or PoE injectors as Power Sourcing Equipment (PSE). They provide a camera or other end devices with 15W (802.3af), 30W (802.3at) or even 45-90W (802.3bt).

Areas of application

Industrial image processing is frequently used in the monitoring of production processes. While in the quality assurance of serial products the output result is often only a simple "pass" or "fail", in other applications of machine vision complex algorithms and methods of artificial intelligence are used to identify and classify properties. But not only in monitoring, also directly in the production process itself, e.g. in robot-controlled "pick-and-place", imaging methods are increasingly replacing classic sensor solutions. Ethernet-based image processing also plays a major role in the area of security and access control, both indoors (supermarkets, warehouses, museums, railway stations) and outdoors (car parks, tunnels, company premises, public places).
Loading...
Upload in progress, please be patient ...