Zebra Technologies Corporation, a leading digital solution provider enabling businesses to intelligently connect data, assets, and people, gives manufacturers the ability to leverage optical character recognition everywhere with its new deep learning software deployed on a range of devices to meet the needs of automotive, pharmaceutical, electronics, and food and beverage manufacturers, as MEPCA found out.
Zebra’s flexible, deep learning optical character recognition (OCR) solution can handle complex use cases, eliminate training time, and ensure stability and ease of use, even for a non-expert. It’s the solution for overcoming the familiar time, cost, training and stability issues of conventional OCR technology.
“Conventional OCR needs a lot of training time, can be unstable when faced with a change in environment, and doesn’t handle complex use cases well,” said Donato Montanari, General Manager and Vice President, Machine Vision at Zebra Technologies. “Many OCR tools require manufacturers to invest a lot of time for something that works in perfect conditions, but too often struggles to read obscure and damaged characters, engraved and embossed formats, characters on reflective and curved surfaces, or perform in changing or harsh lighting conditions.”
Zebra’s deep learning OCR solution is flexible. It can be deployed on desktop PCs, whether Windows, Linux or Linux ARM embedded (ideal for compact devices like Raspberry Pi or Nvidia Jetson), Android handheld devices, and Zebra smart cameras.
Reading identification, compliance, safety and other markings on vehicle tires, test tube label and cap analysis, blood pack labels, and waybill documents for logistics are just some of the use cases that Zebra’s deep learning OCR solution can handle with a level of ease beyond older OCR options.
“This powerful tool delivers very high accuracy straight out-of-the-box and works on both graphics and central processing units,” added Donato. “Manufacturers can also take full control over development and integration with other applications in C++ or .NET using Zebra’s Aurora Vision Library.”
Zebra’s OCR software, powered by deep learning, uses a convolutional neural network that mimics the human brain. It comes pre-trained using thousands of different image samples. This enables the user to create a robust OCR application in just a few simple steps, even if they don’t have machine vision or deep learning expertise.
The simple yet powerful deep learning OCR is easy to use even in complex scenarios. It’s as easy as drawing a box around the characters and letting the tool do the rest. There’s no need to train fonts or maintain libraries. End-users just set the characters’ height, minimum confidence score and match string and are up and running very quickly. Inspections can be rapidly amended on the fly to account for new printing methods or font changes without any of the time-consuming setup required with older OCR technology.
“Workflows are optimised, and the role of the engineer is elevated, as they are taking more and bigger steps toward thinking and acting like data and AI specialists to shape the future of work,” concluded Donato.
Traditional OCR technology has not addressed longstanding challenges faced by manufacturers, and is ill-equipped for the complexity, compliance, speed and volume of today’s manufacturing environments. Zebra’s pioneering the next generation of OCR uses deep learning to deliver a tool that is ready out-of-the-box and can handle complex use cases. As part of Zebra’s deep learning-powered machine vision portfolio, this enables engineers to think and act like data and AI specialists, while also giving programmers and data scientists the tools needed to develop bespoke solutions.