Deep Learning to Drive Machine Vision Segment to $194 Billion in 5 Years, Says ABI

Machine vision technology, riding on deep learning (DL), is advancing and driving mass adoption in automotive, retail, consumer, industrial and surveillance sectors, ABI Research reported Wednesday. It forecasts machine vision technology will grow at a 53 percent annual compound growth rate through 2023 to $193.8 billion generated from services and hardware. Machine vision vendors previously relied on hard-coded feature detection techniques applied in highly controlled environments, such as detecting a particular object on a production line, ABI said, but DL-based systems are more flexible, capable of recognizing many object types and deployable in a range of circumstances. The researcher gave as examples cashier-less stores such as Amazon Go, where cameras track movements of both customers and items in the store, or in autonomous driving where systems can make distinctions between multiple types of vehicles and drivers. Roughly 37.8 million vehicles shipped in 2023 will contain between level 2 to 5 advanced driver-assistance systems (ADAS). More than half of the 34.4 million level 2 ADAS systems shipped that year will use DL-based machine vision, and level 3-5 vehicles will all use the approach, it said.