How Machine Vision Broadens the Spectrum of Embedded Systems?

By Embedded Advisor | Tuesday, November 19, 2019

Machine vision is all set to enhance the potential applications of embedded systems via component-based imaging systems.

FREMONT, CA: Machine vision is increasingly contributing to the world of embedded systems. The need for efficiency, coupled with the pressure to restrain the cost, is leading to component-based imaging systems. With the advancement in technology and an increase in the number of smart industrial wearables, the need to lace the devices with imaging capabilities has also gained essence.

Embedded vision systems have been largely used in mobile devices. However, advancement in technology is expanding the possible use cases of embedded vision into other landscapes as well. In the coming few years, embedded vision technology will be found almost everywhere, ranging from heavily automated smart factories to ubiquitous everyday devices.

Although a PC-based vision system offers good performance, it can be bulky and complex. On the other hand, embedded vision systems consist of an independent computer system that can be integrated directly into an electrical system or a larger mechanical system. Further, embedded systems are cheaper and easier to use than the PC-based counterparts. Minimal maintenance overheads also make embedded vision systems a lucrative option.

In the embedded world, a camera integration works with a USB or GigE interface, which is analogous to a plug-and-play solution connected to a PC. Camera manufacturers are providing their software development kit (SDK) in a format that also works on an ARM platform so that users can reconcile a camera in a similar fashion as on a personal computer (PC).

An embedded system can be specialized to an even higher level for certain applications. System on Module (SoM), which is a board-level circuit, integrates system functions in a single module. The highly compact SoM modules can contain only a processor such as a memory chip, microcontrollers, or other essential components. Computer vision capability will enable SoM systems to perform image processing, thereby reducing latency and bandwidth consumption.

Machine vision is improving the efficiency and throughput of the embedded systems. With powerful processing capabilities, embedded vision systems will contribute immensely to a wide range of industrial applications.