The system receives video footage from six cameras installed in the display cabinet and automatically detects goods using deep learning technology to ring up purchase. (image: Inha University)
INCHEON, Feb. 26 (Korea Bizwire) — A South Korean research team has developed technology that allows for an unmanned checkout system.
Researchers led by Prof. Song Byung-cheol from Inha University, in collaboration with Interpark, a major South Korean online shopping mall, developed a computer vision technology for smart display cabinets.
The system receives video footage from six cameras installed in the display cabinet and automatically detects goods using deep learning technology to ring up purchase.
The system was known to detect up to 95 percent of all randomly selected goods to proceed to checkout.
The research team said that the new system can distinguish multiple motions at once, which resolves the issue of not being able to detect items when a customer shuffles them.
“The new algorithm can be used on ordinary computers, which raises its price competitiveness,” said Song. “We will continue to develop technologies that can deal with more complex form of transactions.”
Kevin Lee (kevinlee@koreabizwire.com)
