Optical character recognition (OCR) is fast becoming a key part of manufacturing and packaging quality assurance, as more companies use advanced cameras and sensors to recognize text characters and information automatically and verify correct parts, assemblies, finished goods, labels and product packaging.
However, as OCR technology has made its way into production and packaging processes, one of the biggest challenges has been training cameras and sensors to read, recognize and extract the right characters, especially when dealing with low-contrast, poor lighting and confusing backgrounds. These conditions can make it hard for cameras and sensors to read characters, and even if they can, training them to do so is often a time-consuming task.
Deep-Learning Optical Character Recognition Revolution
But now there has been a big breakthrough in OCR technology with Zebra’s introduction of its machine vision sensors and Aurora Focus software. These solutions use Zebra’s Deep-Learning OCR to automatically extract text and information from images without text training. They automatically adapt to different fonts, characters, contrast, lighting and backgrounds, and they provide extremely reliable reads with machine accuracy and precision.
Instead of having to train and retrain your cameras or sensors to recognize different fonts or text, you simply use Zebra’s Aurora Focus software to draw a box around an image that your sensor has captured, and you tell the software to read, recognize and extract the characters. Zebra’s software then teaches its own algorithms, and you can set thresholds to help it differentiate between different character strings such as mathematical or alphanumeric characters, and to account for different sizes or spacing.
Sensors and Software Teach Themselves
Since Zebra’s sensors and software train themselves, they learn and adapt almost instantly, and they can start reading and extracting the right text and information in just a few minutes. Zebra’s sensor captures the image, and the software and algorithms do the rest. They instantly recognize and extract the right text and information, even at high-speeds, from components, finished goods, labels or packages, even if they’re moving rapidly on conveyors.
This faster and simpler approach to OCR means you won’t end up spending hours, days or weeks trying to get your cameras or sensors to recognize a new font, capture a letter, or character that’s skewed, rotated, stretched or compressed. The sensors and software teach themselves, so you don’t have to train them.
Additionally, the initial OCR setup with Zebra is extremely simple, even for a beginner who has never worked with industrial automation sensors. Zebra’s Aurora Focus software provides a truly modern and streamlined user experience, with simple and automated tools, radio buttons, sliders and drop-down menus to define every job and adjust every setting. It also includes built-in videos and tutorials to walk you through all of its machine vision and fixed scanning capabilities, so you can actually learn as you go.
Single Platform
Zebra’s machine vision sensors also offer a massive array of additional capabilities, far beyond OCR. Its plug-and-play sensors provide machine vision and fixed industrial scanning capabilities from a single hardware and software platform, so you don’t need to buy and learn different smart cameras and sensors from different companies to get the results you need.
Instead, you can use the same sensors to scan barcodes, capture DPMs, perform sophisticated visual analysis and automate everything from manufacturing traceability to routing, counting and sorting; label and packaging checks; and visual inspections.
To cover this wide range of applications and use cases, Zebra’s sensors and software offer a full complement of powerful tools, including the following capabilities:
1D/2D barcode and DPM scanning
Object location
Pixel counting
Brightness detection
Contrast detection
Edge detection
Distance measurement
Pattern recognition
Caliper tools
Circle finding
Filters
OCV/flaw recognition
All of these capabilities are available from one hardware and software platform, so you don’t need to buy separate devices or software applications. You also don’t have to go through multiple learning curves to get the results you need.
With Zebra’s sensors and software, everything has been consolidated and simplified, so it’s faster, less complicated, and less expensive to get your industrial automation jobs up and running. Most of all, though, you can count on extremely reliable results, which will help you ensure the highest quality in your manufacturing and packaging.
Optical character recognition (OCR) is fast becoming a key part of manufacturing and packaging quality assurance, as more companies use advanced cameras and sensors to recognize text characters and information automatically and verify correct parts, assemblies, finished goods, labels and product packaging.
However, as OCR technology has made its way into production and packaging processes, one of the biggest challenges has been training cameras and sensors to read, recognize and extract the right characters, especially when dealing with low-contrast, poor lighting and confusing backgrounds. These conditions can make it hard for cameras and sensors to read characters, and even if they can, training them to do so is often a time-consuming task.
Deep-Learning Optical Character Recognition Revolution
But now there has been a big breakthrough in OCR technology with Zebra’s introduction of its machine vision sensors and Aurora Focus software. These solutions use Zebra’s Deep-Learning OCR to automatically extract text and information from images without text training. They automatically adapt to different fonts, characters, contrast, lighting and backgrounds, and they provide extremely reliable reads with machine accuracy and precision.
Instead of having to train and retrain your cameras or sensors to recognize different fonts or text, you simply use Zebra’s Aurora Focus software to draw a box around an image that your sensor has captured, and you tell the software to read, recognize and extract the characters. Zebra’s software then teaches its own algorithms, and you can set thresholds to help it differentiate between different character strings such as mathematical or alphanumeric characters, and to account for different sizes or spacing.
Sensors and Software Teach Themselves
Since Zebra’s sensors and software train themselves, they learn and adapt almost instantly, and they can start reading and extracting the right text and information in just a few minutes. Zebra’s sensor captures the image, and the software and algorithms do the rest. They instantly recognize and extract the right text and information, even at high-speeds, from components, finished goods, labels or packages, even if they’re moving rapidly on conveyors.
This faster and simpler approach to OCR means you won’t end up spending hours, days or weeks trying to get your cameras or sensors to recognize a new font, capture a letter, or character that’s skewed, rotated, stretched or compressed. The sensors and software teach themselves, so you don’t have to train them.
Additionally, the initial OCR setup with Zebra is extremely simple, even for a beginner who has never worked with industrial automation sensors. Zebra’s Aurora Focus software provides a truly modern and streamlined user experience, with simple and automated tools, radio buttons, sliders and drop-down menus to define every job and adjust every setting. It also includes built-in videos and tutorials to walk you through all of its machine vision and fixed scanning capabilities, so you can actually learn as you go.
Single Platform
Zebra’s machine vision sensors also offer a massive array of additional capabilities, far beyond OCR. Its plug-and-play sensors provide machine vision and fixed industrial scanning capabilities from a single hardware and software platform, so you don’t need to buy and learn different smart cameras and sensors from different companies to get the results you need.
Instead, you can use the same sensors to scan barcodes, capture DPMs, perform sophisticated visual analysis and automate everything from manufacturing traceability to routing, counting and sorting; label and packaging checks; and visual inspections.
To cover this wide range of applications and use cases, Zebra’s sensors and software offer a full complement of powerful tools, including the following capabilities:
All of these capabilities are available from one hardware and software platform, so you don’t need to buy separate devices or software applications. You also don’t have to go through multiple learning curves to get the results you need.
With Zebra’s sensors and software, everything has been consolidated and simplified, so it’s faster, less complicated, and less expensive to get your industrial automation jobs up and running. Most of all, though, you can count on extremely reliable results, which will help you ensure the highest quality in your manufacturing and packaging.
To learn more about Zebra machine vision and its Deep-Learning OCR, visit our Zebra industrial automation.
Recent Posts