The latest supermarket scanner developed by Toshiba Tec may make conventional barcodes in supermarkets obsolete.

Currently, at most supermarkets, a laser scanner is used to recognize barcodes, but the newly developed Object Recognition Scanner instantly recognizes the merchandise itself, be it fresh produce or packaged goods, by identifying it directly via a camera.

“Fruit and vegetables in supermarkets don’t usually have barcodes, because they’re put out while they’re fresh. So these items can’t be read at the register using barcodes, which means staff need to input data to record them. If staff are part-time employees, they may not recognize some items, which can cause delays. We’re developing this new scanner to solve that problem.”

This scanner utilizes pattern recognition technology being developed by Toshiba. By recognizing the difference between merchandise and other objects and discarding the latter as noise, the scanner can operate at high speeds.

“As you can actually see in the picture, the surroundings are totally dark. Ordinarily, with a camera, the surroundings would appear as well, but here, only the subjects like apples appear. That means the noise is eliminated from the beginning, so the scanner can recognize objects very fast, even if they’re moving. This technology is the key to the new system.”

“In this demo, there are three kinds of apples: Fuji, Jonagold, and Mutsu. The Fuji and Jonagold originally come from the same stock, so if you’re not really familiar with apples, they might look the same. But this scanner can distinguish them, by recognizing subtle differences in pattern and coloration.”

As well as fruits and vegetables, the scanner can recognize printed items, such as beer cans and coupons.

“Getting the scanner to learn items in a store isn’t practical, so we’ll ship it with a database of agricultural produce and other items as needed. It takes a year for all vegetables to appear in their seasons, so for a machine to read them, we have to create a database. So that anyone can use the scanner without needing to practice, it has to be able to read items consistently, whether they’re touching it, far away from it, or moving. Our main task in development from now on is tuning the scanner so it can do that.”

This Video is provided to you by DigInfo.tv, AkihabaraNews Official Partner.