Researchers are predicting considerable growth in embedded vision technology, particularly in the automotive, industrial automation, physical security, and business intelligence markets. Shipments of embedded vision devices for use in those markets are expected to exceed 14 million units by 2018, up from nearly four million units this year, according to a recent report from researcher IHS Technology.
Embedded vision uses a combination of embedded systems and computer vision technology. It allows devices to use video inputs to better understand their environment, applying logic and decision-making to video signals, the researcher explains, adding that the maturity of embedded vision algorithms varies by market.
“For instance, while embedded vision technology has been active for some time in markets like physical security and industrial automation, the consumer industry represents more of an emerging opportunity,” according to the IHS report.
“However, despite the synergies in algorithm requirements across application markets, there are very few vendors that are active across multiple applications,” the report also explains. “In some markets, like automotive, the long sales cycles and high qualifying requirements have limited new competition. In others, such as physical security, the fragmented equipment market means that algorithms need to be optimized for a large number of products, which can act as a barrier to new entrants.”
Demand for embedded vision technology in automotive markets in particular will spur rapid growth in the decade ahead, IHS adds.
Recent moves by large distributors of electronic components and technology solutions underscore that forecast. Avnet Electronics Marketing, for instance, has made recent announcements about its ability to sell and support the technology. Earlier this spring, Avnet was promoting Texas Instruments’ machine vision technology, which is designed “to be the eyes of your automobile, able to see, interpret, warn, and eventually take action to safeguard the car’s occupants,” according to a TI online technical article.
“The machine vision system gives the car the situational awareness of what is going on around the car using ultrasound, radar, cameras or all three at the same time” said Hannes Estl, a marketing engineer in TI systems marketing. “The camera systems will use the information for the car to act without the driver. For example, instead of a ‘lane departure warning’ system, it becomes ‘lane keep assist’ system, where the car detects that the driver is about to leave their lane inadvertently and slowly auto-steers the car back into its lane.”
Other potential applications for the technology include emergency braking, Hannes said. And although estimates vary, many industry experts predict that driverless cars will be rolling off assembly lines sometime in the next decade.
“… By 2025, it is believed there will be fully autonomous driving where you can actually lean back and read a book or watch a show,” said Hannes. “It is very exciting to be working on something like this that has a huge impact on how we drive and how we perceive the car.”
Avnet expands on embedded vision technology with its own article about Getting Started in Automotive Smart Vision Design.
The distributor also announced the release of its MicroZED Embedded Vision Carrier Card kit designed to help engineers more quickly and efficiently deploy solutions that incorporate the MicroZed System-on-Module. The kit builds on the MicroZed SOM by providing a video-specific carrier card with camera connector for optional camera modules featuring image sensors from ON Semiconductor and Toshiba America Electronic Components.
Two versions of the kit are available: the MicroZed™ Embedded Vision Carrier Card Kit, without a MicroZed SOM, for users who already have a MicroZed SOM; and the MicroZed™ Embedded Vision Development Kit, including a MicroZed 7020 SOM and voucher for the Vivado® Design Edition design suite, device-locked to the Zynq 7020.
